交通小区划分问题的整数规划建模与优化算法研究
发布时间：2019-06-25 00:21
【摘要】：城市交通是城市发展和进步的决定性因素,与社会经济的发展相辅相成。交通超前于社会经济的发展,是发达国家保持经济长期快速发展的一个重要因素,反之交通问题则会成为制约经济发展的主要瓶颈。为此,如何做好交通规划是我国当前急需解决的迫切问题,也是保障未来发展不可忽视的重要问题。交通调查作为交通规划的主要内容之一,是获取交通数据的基本方法和必要手段。通过交通调查获得的数据是未来需求预测、规划方案制定的主要依据,其真实性和可靠性直接决定着规划结果。由于交通源的数量巨大,交通调查一般以交通小区作为空间统计单元,要求确定合理的统计单元,不能过细而造成资源浪费,又不能过粗而影响后续的规划结果,这就要求使用科学的方法制定这些统计单元。有鉴于此,本文以城市交通规划为研究背景,运用整数规划建模求解理论与方法,分别针对交通小区划分问题的数据抽象和最优化建模、空间邻接约束的建模、分区个数已知的交通小区划分问题(基本交通小区划分问题)以及考虑分区个数决策的交通小区划分问题(扩展的交通小区划分问题)的建模与求解开展了系统的研究,主要研究工作如下文所述。(1)围绕交通规划中交通小区的实际划分方法、理论研究中最优交通小区划分问题的数据抽象过程、优化目标的选择与建立以及优化框架与求解方法等几个方面,开展了较全面的综述研究工作,并在此基础上针对空间单元的邻接矩阵构造方法、空间单元的可达性计算方法以及基于可达性的交通小区划分的优化目标建立等问题开展了算法设计与建模工作。该项研究工作的具体内容包括：针对交通活动实践中的交通小区,以文献资料为依据阐述了优化划分交通小区的意义,并概述了交通规划管理实践中交通小区划分的原则及方法；从微观角度,针对交通小区划分问题的理论研究,重点围绕使用优化技术解决交通小区划分问题时设立何种优化目标以及如何抽象构建优化目标的问题,分别针对建立优化目标前对连续的空间研究区域进行离散化的方法、离散化后面状空间单元的可达性度量方法、现实与文献中划分交通小区最优目标的内容与形式以及基于可达性构建的交通小区划分最优目标的内容与表达形式四个方面,开展了文献综述工作；在此基础上分别提出了基于Z型顺序编码的基本地理单元邻接矩阵的构造算法、使用路段局部深度值度量基本地理单元可达性的方法以及两个交通小区划分优化目标的数学表达形式—最小化总区内加权出行费用目标和最小化总区内可达性差异目标；从宏观角度,针对交通小区划分问题研究中的理论优化框架与方法,综述了最优交通小区划分问题的相关研究成果,将其优化问题的框架概括为分区依据数据、最优目标、基本分区约束、问题约束、算法以及评价分区解的指标这五个要素；并分别从不同类型的分区依据数据和求解方法两个方面对交通小区划分问题的理论研究进行了进一步的综述。(2)围绕使用整数规划技术对最优交通小区空间邻接约束建模的方法,针对文献中保障最优交通小区分区解满足空间邻接约束的方法进行了简要的综述,在此基础上,针对空间邻接约束整数规划建模方法以及不同建模方法对问题求解效率的影响这两个方面的问题开展了建模求解与算例分析等研究工作。该项研究工作的具体内容包括：在综述最优交通小区空间邻接约束建模的方法的基础上,修正文献中邻接约束整数规划建模的方法以适用于交通小区划分问题,给出了最小生成树表示、顺序路径表示以及网络流表示三种对空间邻接约束建模的方法；提出了基于邻接矩阵表示的保障分区一阶邻接约束的一种新建模方法；分别从模型的决策变量规模与模型的约束规模两个角度,给出了四种模型的变量、约束数量的推算过程以及理论计算表达式；基于对四种模型决策变量与约束规模的理论值估算,讨论并比较与分析了四种模型的求解复杂度；基于两个小规模仿真算例对四种模型的求解过程及原理、求解效率进行了比较与分析,并简要探讨了本文所提出的基于邻接矩阵的建模方法的适用范围。(3)围绕基本交通小区划分问题,研究了在分区个数已知的情景下,基于P中位问题模型的交通小区划分问题的0-1整数规划建模及求解方法。该项研究工作的具体内容包括：根据P中位问题模型,在假设分区决策参数已知的条件下,建立了以最小异质性为优化目标,满足空间邻接约束的整数规划模型的基本形式—最优交通小区划分问题的P中位问题模型(TAZ-PMP模型),并分析与探讨了模型的性质；将空间邻接约束与TAZ-PMP模型分离,由此将最优交通小区划分问题分解为构造列池以及求解最优下料问题两个过程,基于此,将邻接约束作为隐枚举规则,给出了通过构造TAZ-PMP模型主问题的初始化列池进而将TAZ-PMP模型转化为求解最优下料问题的精确求解算法(TAZ-IE算法)；使用拉格朗日替代松弛技术对TAZ-PMP模型进行分解,结合近似求解拉格朗日对偶问题的搜索算法、优化求解拉格朗日对偶问题的下降梯度算法以及TAZH算法给出了基于拉格朗日替代松弛的局部搜索启发式算法(TAZ-LSLSH算法)；通过识别、推导拉格朗日替代松弛和下降花费问题中的共同优化的子问题,将拉格朗日替代松弛与列生成过程相结合,给出了基于拉格朗日替代松弛方法的主问题与价格子问题的分解过程,并基于此设计了基于拉格朗日替代松弛技术的列生成算法(TAZ-LSCG算法)：使用基于OR-Library和Pcb3038修改的仿真数值算例对三种算法的求解过程及求解效率进行了比较与分析。(4)对基本交通小区划分问题进行扩展,围绕考虑分区个数未知的情景下,交通小区划分问题的非线性混合整数规划建模、模型的解析求解特征以及模型的求解算法等内容,开展了建模、算法设计与算例分析等研究工作。该项研究工作的具体内容包括：考虑分区组成、分区中心设置以及分区个数三个决策变量,建立了以最小地理误差为目标,以各分区空间可达性同质、各分区面积同质为主要约束条件的最优交通小区划分问题的混合整数规划模型(K-TAZ模型)；在分析了基本地理单元设置与最优分区问题之间关系的基础上给出了影响区的定义,给出了构造影响区的三个核心要素为邻接关系谓词、同质性度量的划分谓词以及影响区构成规则,设计了构造影响区的启发式搜索算法；将最优交通小区划分问题的求解空间离散单元由基本地理单元改为影响区,对提出的K-TAZ模型进行了改进,使最优分区问题更容易求解；对K-TAZ模型进行了解析推导,给出了关于模型最优分区个数解下界的引理与定理,以缩小分区个数取值范围为目标,提出了确定最大分区个数下界的方法；基于K-TAZ模型中与分区个数决策变量相关的同质性约束,作为搜索可行分区个数的限制条件,结合引理给出了可行分区个数的隐枚举算法；分析了K-TAZ模型无解情景以及导致K-TAZ模型无解的理论原因；设计了包含两个阶段的聚合式聚类启发式算法以重构求解空间使K-TAZ模型可解；通过对比仅包含一阶段的聚合式聚类启发过程与所提出算法的启发式过程中求解空间结构的变化,说明了所提出算法的有效性；将最大Kmin域缩减方法与约束规划模型求解过程结合,给出了求解K-TAZ模型的约束规划方法；将可行分区个数隐枚举域缩减方法与P中位问题模型求解过程结合,给出了求解K-TAZ模型的P中位模型方法；使用基于Pcb3038修改的仿真数值算例对两种算法的求解过程及求解效率进行了比较与分析。(5)结合苏州工业园区公交规划中最优交通小区划分的实际案例,开展了抽象建模与算例分析等应用研究。该项研究工作的具体内容包括：基于本文提出的抽象建模与计算方法,给出了案例的数据模型及其建立过程；考虑分区个数固定的情景,将案例的最优交通小区划分问题抽象为最小化分区异质性的TAZ-PMP模型；分别使用IE算法、LSLSH算法以及LSCG算法给出了案例最小化分区异质性TAZ-PMP模型的最优交通小区划分方案,并对算法求解实例的效率进行了讨论与分析；考虑分区个数未知的情景,将案例的最优交通小区划分问题抽象为最小化地理误差的K-TAZ模型；分别使用CP方法以及PMP方法给出了最小化地理误差K-TAZ模型的最优交通小区划分方案,并对算法求解实例的效率进行了相关的讨论与分析。
[Abstract]:Urban traffic is the decisive factor of urban development and progress, and is mutually reinforcing with the development of social economy. The development of the traffic ahead of the social economy is an important factor for the developed countries to maintain the long-term and fast development of the economy. To this end, how to do well the traffic planning is the urgent problem to be solved urgently in our country, and it is also an important issue to guarantee the future development. As one of the main contents of the traffic planning, the traffic survey is the basic method and necessary means for obtaining the traffic data. The data obtained through the traffic investigation is the main basis for future demand forecast and planning, and its authenticity and reliability directly determine the planning result. because the number of the traffic sources is large, the traffic survey is generally used as a space statistical unit in a traffic cell, a reasonable statistical unit is required to be determined, the resource waste can not be caused to be too thin, and the subsequent planning result can not be influenced by rough, This requires the use of scientific methods for the development of these statistical units. In view of the above, this paper uses the urban traffic planning as the research background, and uses the integer programming modeling to solve the theory and the method, respectively the data abstraction and optimization modeling of the traffic cell division problem, the modeling of the space adjacency constraint, The model and solution of the traffic cell division problem (the basic traffic cell division problem) and the traffic cell division problem (extended traffic cell division problem) considering the number decision of the division are studied in this paper. The main research work is as follows. (1) The comprehensive review and research work is carried out in the aspects of the actual division method of the traffic cell in the traffic planning, the data abstraction process of the optimal traffic cell division problem in the theoretical study, the selection and establishment of the optimization target, the optimization framework and the solution method, and the like. On the basis of this, the algorithm design and modeling of the space unit's neighbor matrix construction method, the reachability calculation method of the space unit and the optimization target establishment based on the accessibility of the traffic cell are carried out. The concrete contents of this research include: for the traffic district in the practice of traffic activity, the paper expounds the significance of the optimization of the traffic district based on the literature, and outlines the principle and method of the traffic cell division in the traffic planning management practice; from the micro angle, In order to solve the problem of traffic cell division, this paper focuses on how to use the optimization technique to solve the problem of the optimization goal and how to abstract and construct the optimization target. a method for discretizing a continuous space research area before an optimization target is established, and a reachability measurement method for the discretized back space unit, The paper reviews the content and form of the optimal target of the traffic cell and the content and the expression form of the optimal target for the traffic subdistrict based on the accessibility. On the basis of this, the structure algorithm of the adjacent matrix of the basic geographical unit based on the Z-type sequence coding is presented, the method for measuring the accessibility of the basic geographic unit by using the local depth value of the road section and the mathematical expression form of the two traffic cell division optimization targets are used to minimize the weighted travel expense target in the total area and to minimize the reachability difference target in the total area; and from the macroscopic angle, In view of the theory optimization framework and method in the research of traffic cell division, the relevant research results of the optimal traffic cell division problem are summarized, and the frame of the optimization problem is summarized as the partition basis data, the optimal objective, the basic partition constraint, the problem constraint, The five elements of the algorithm and the index for evaluating the partition solution are summarized, and the theoretical research on the division of traffic cells from different types of partitions according to the data and the method of solution is further reviewed. and (2) a method for modeling the spatial adjacent constraint of the optimal traffic cell by using an integer programming technique, In order to solve the problems of space-adjacent constrained integer programming modeling and different modeling methods, the problems in the two aspects of the problem-solving efficiency are studied, such as the modeling solution and the numerical example analysis. The specific content of the research work includes: on the basis of summarizing the method of the optimal traffic cell space adjacency constraint modeling, the method of the adjacent constrained integer programming modeling in the document is applied to the traffic cell division problem, and the minimum spanning tree representation is given, The sequential path representation and the network flow represent three methods for modeling the spatial adjacency constraint; a new modeling method for guaranteeing the first order adjacency constraint of the partition based on the adjacency matrix representation is proposed; and the two angles of the scale of the decision variables of the model and the constraint scale of the model are respectively obtained, In this paper, four model variables, the calculation process of the constraint number and the theoretical calculation expression are given, and the solution complexity of the four models is discussed and compared based on the theoretical value estimation of the four model decision variables and the constraint scale. Based on two small-scale simulation examples, the solution process and principle of four models are compared and analyzed, and the application scope of the method based on the adjacency matrix is briefly discussed. (3) The problem of the division of the basic traffic is studied, and the 0-1 integer programming model and the method for solving the traffic cell division problem based on the P-bit problem model are studied. The specific contents of the study include: according to the P-position problem model, under the assumption that the partition decision parameter is known, the minimum heterogeneity is established as the optimization target, The basic form of an integer programming model satisfying the spatial adjacency constraint is the P center problem model (TAZ-PMP model) of the optimal traffic cell division problem, and the property of the model is analyzed and discussed; and the space adjacency constraint is separated from the TAZ-PMP model, in that method, the problem of the optimal traffic cell division is decomposed into two processes of constructing a column pool and solving the optimal blanking problem, The TAZ-PMP model is transformed into an accurate solution algorithm (TAZ-IE) for solving the optimal blanking problem by constructing the initialization column pool of the main problem of the TAZ-PMP model, and the TAZ-PMP model is decomposed using the Lagrange substitution relaxation technique, and the search algorithm for solving the Lagrange dual problem is combined with the approximate solution, In this paper, a local search heuristic algorithm (TAZ-LSLSH algorithm) based on the Lagrange's alternative relaxation is given in this paper. In this paper, the decomposition process of the main problem and the price subproblem based on the Lagrange's alternative relaxation method is given by the combination of the Lagrange's replacement relaxation and the column generation process, and the column generation algorithm (TAZ-LSCG algorithm) based on the Lagrange substitution relaxation technique is designed. A numerical example based on OR-Library and Pcb3038 is used to compare and analyze the process and efficiency of the three algorithms. (4) the basic traffic cell division problem is expanded, and the modeling is carried out around the content of the non-linear mixed integer programming modeling of the traffic cell division problem, the analysis and solving characteristics of the model and the solution algorithm of the model, The design of the algorithm and the analysis of the numerical example and so on. The specific contents of the study include: considering the composition of the partition, the setting of the center of the partition and the three decision variables of the number of the zones, the objective of the minimum geographical error is established, and the spatial accessibility of each partition is homogeneous, the mixed integer programming model (k-taz model) of the optimal traffic cell division problem with the uniform partition area as the main constraint condition is given; the definition of the influence area is given on the basis of analyzing the relation between the basic geographic unit setting and the optimal partition problem, The three core elements of the structure influence area are given as the adjacent relation predicate, the partition predicate of the homogeneity measure and the influence area constitute the rule, and the heuristic search algorithm for constructing the influence area is designed; The method for solving the problem of optimal traffic cell division is changed from the basic geographic unit to the influence area, the proposed K-TAZ model is improved, the optimal partition problem is more easily solved, and the K-TAZ model is analyzed and derived, In this paper, the lemma and the theorem of the lower bound of the number of the optimal partition of the model are given, and the method of determining the lower bound of the maximum number of partitions is put forward. Based on the homogeneity constraint associated with the number of the partition number in the K-TAZ model, In this paper, the implicit enumeration algorithm for the number of feasible partitions is given as the limit of the number of feasible partitions, and the non-solution scenarios of the K-TAZ model and the theoretical reasons leading to the non-solution of the K-TAZ model are analyzed. In this paper, a heuristic algorithm with two stages of aggregation is designed to reconstruct the solution space so that the K-TAZ model can be solved, and the validity of the proposed algorithm is explained by comparing the heuristic process of the aggregation-type clustering with only one stage and the heuristic process of the proposed algorithm. In this paper, the maximum Kmin domain reduction method is combined with the constraint planning model solving process, and a constraint planning method for solving the K-TAZ model is given, and the method for solving the K-TAZ model is given by combining the feasible partition number of the hidden enumeration domain reduction method with the P median problem model solving process, and the method for solving the P-type model of the K-TAZ model is provided. The solution process and efficiency of two algorithms are compared and analyzed using a simulation numerical example based on Pcb3038 modification. (5) In combination with the actual case of the optimal traffic cell division in the public transport planning of the Suzhou Industrial Park, the application research of the abstract modeling and an example analysis is carried out. The concrete content of this research includes: based on the abstract modeling and calculation method presented in this paper, the data model of the case and the process of establishing the case are given; considering the fixed situation of the number of the partitions, the optimal traffic cell division problem of the case is abstracted into the TAZ-PMP model which minimizes the partition heterogeneity; By using the IE algorithm, the LSLSH algorithm and the LSCG algorithm, the optimal traffic cell division scheme for the case-minimized partition heterogeneity TAZ-PMP model is given, and the efficiency of the algorithm is discussed and analyzed. The optimal traffic cell division problem of the case is abstracted as the K-TAZ model of minimizing the geographical error; the optimal traffic cell division scheme of minimizing the geographic error K-TAZ model is given by using the CP method and the PMP method, and the efficiency of the algorithm solution case is discussed and analyzed.
【学位授予单位】：东北大学
【学位级别】：博士
【学位授予年份】：2014
【分类号】：U491.12;TP301.6
本文编号：2505495
[Abstract]:Urban traffic is the decisive factor of urban development and progress, and is mutually reinforcing with the development of social economy. The development of the traffic ahead of the social economy is an important factor for the developed countries to maintain the long-term and fast development of the economy. To this end, how to do well the traffic planning is the urgent problem to be solved urgently in our country, and it is also an important issue to guarantee the future development. As one of the main contents of the traffic planning, the traffic survey is the basic method and necessary means for obtaining the traffic data. The data obtained through the traffic investigation is the main basis for future demand forecast and planning, and its authenticity and reliability directly determine the planning result. because the number of the traffic sources is large, the traffic survey is generally used as a space statistical unit in a traffic cell, a reasonable statistical unit is required to be determined, the resource waste can not be caused to be too thin, and the subsequent planning result can not be influenced by rough, This requires the use of scientific methods for the development of these statistical units. In view of the above, this paper uses the urban traffic planning as the research background, and uses the integer programming modeling to solve the theory and the method, respectively the data abstraction and optimization modeling of the traffic cell division problem, the modeling of the space adjacency constraint, The model and solution of the traffic cell division problem (the basic traffic cell division problem) and the traffic cell division problem (extended traffic cell division problem) considering the number decision of the division are studied in this paper. The main research work is as follows. (1) The comprehensive review and research work is carried out in the aspects of the actual division method of the traffic cell in the traffic planning, the data abstraction process of the optimal traffic cell division problem in the theoretical study, the selection and establishment of the optimization target, the optimization framework and the solution method, and the like. On the basis of this, the algorithm design and modeling of the space unit's neighbor matrix construction method, the reachability calculation method of the space unit and the optimization target establishment based on the accessibility of the traffic cell are carried out. The concrete contents of this research include: for the traffic district in the practice of traffic activity, the paper expounds the significance of the optimization of the traffic district based on the literature, and outlines the principle and method of the traffic cell division in the traffic planning management practice; from the micro angle, In order to solve the problem of traffic cell division, this paper focuses on how to use the optimization technique to solve the problem of the optimization goal and how to abstract and construct the optimization target. a method for discretizing a continuous space research area before an optimization target is established, and a reachability measurement method for the discretized back space unit, The paper reviews the content and form of the optimal target of the traffic cell and the content and the expression form of the optimal target for the traffic subdistrict based on the accessibility. On the basis of this, the structure algorithm of the adjacent matrix of the basic geographical unit based on the Z-type sequence coding is presented, the method for measuring the accessibility of the basic geographic unit by using the local depth value of the road section and the mathematical expression form of the two traffic cell division optimization targets are used to minimize the weighted travel expense target in the total area and to minimize the reachability difference target in the total area; and from the macroscopic angle, In view of the theory optimization framework and method in the research of traffic cell division, the relevant research results of the optimal traffic cell division problem are summarized, and the frame of the optimization problem is summarized as the partition basis data, the optimal objective, the basic partition constraint, the problem constraint, The five elements of the algorithm and the index for evaluating the partition solution are summarized, and the theoretical research on the division of traffic cells from different types of partitions according to the data and the method of solution is further reviewed. and (2) a method for modeling the spatial adjacent constraint of the optimal traffic cell by using an integer programming technique, In order to solve the problems of space-adjacent constrained integer programming modeling and different modeling methods, the problems in the two aspects of the problem-solving efficiency are studied, such as the modeling solution and the numerical example analysis. The specific content of the research work includes: on the basis of summarizing the method of the optimal traffic cell space adjacency constraint modeling, the method of the adjacent constrained integer programming modeling in the document is applied to the traffic cell division problem, and the minimum spanning tree representation is given, The sequential path representation and the network flow represent three methods for modeling the spatial adjacency constraint; a new modeling method for guaranteeing the first order adjacency constraint of the partition based on the adjacency matrix representation is proposed; and the two angles of the scale of the decision variables of the model and the constraint scale of the model are respectively obtained, In this paper, four model variables, the calculation process of the constraint number and the theoretical calculation expression are given, and the solution complexity of the four models is discussed and compared based on the theoretical value estimation of the four model decision variables and the constraint scale. Based on two small-scale simulation examples, the solution process and principle of four models are compared and analyzed, and the application scope of the method based on the adjacency matrix is briefly discussed. (3) The problem of the division of the basic traffic is studied, and the 0-1 integer programming model and the method for solving the traffic cell division problem based on the P-bit problem model are studied. The specific contents of the study include: according to the P-position problem model, under the assumption that the partition decision parameter is known, the minimum heterogeneity is established as the optimization target, The basic form of an integer programming model satisfying the spatial adjacency constraint is the P center problem model (TAZ-PMP model) of the optimal traffic cell division problem, and the property of the model is analyzed and discussed; and the space adjacency constraint is separated from the TAZ-PMP model, in that method, the problem of the optimal traffic cell division is decomposed into two processes of constructing a column pool and solving the optimal blanking problem, The TAZ-PMP model is transformed into an accurate solution algorithm (TAZ-IE) for solving the optimal blanking problem by constructing the initialization column pool of the main problem of the TAZ-PMP model, and the TAZ-PMP model is decomposed using the Lagrange substitution relaxation technique, and the search algorithm for solving the Lagrange dual problem is combined with the approximate solution, In this paper, a local search heuristic algorithm (TAZ-LSLSH algorithm) based on the Lagrange's alternative relaxation is given in this paper. In this paper, the decomposition process of the main problem and the price subproblem based on the Lagrange's alternative relaxation method is given by the combination of the Lagrange's replacement relaxation and the column generation process, and the column generation algorithm (TAZ-LSCG algorithm) based on the Lagrange substitution relaxation technique is designed. A numerical example based on OR-Library and Pcb3038 is used to compare and analyze the process and efficiency of the three algorithms. (4) the basic traffic cell division problem is expanded, and the modeling is carried out around the content of the non-linear mixed integer programming modeling of the traffic cell division problem, the analysis and solving characteristics of the model and the solution algorithm of the model, The design of the algorithm and the analysis of the numerical example and so on. The specific contents of the study include: considering the composition of the partition, the setting of the center of the partition and the three decision variables of the number of the zones, the objective of the minimum geographical error is established, and the spatial accessibility of each partition is homogeneous, the mixed integer programming model (k-taz model) of the optimal traffic cell division problem with the uniform partition area as the main constraint condition is given; the definition of the influence area is given on the basis of analyzing the relation between the basic geographic unit setting and the optimal partition problem, The three core elements of the structure influence area are given as the adjacent relation predicate, the partition predicate of the homogeneity measure and the influence area constitute the rule, and the heuristic search algorithm for constructing the influence area is designed; The method for solving the problem of optimal traffic cell division is changed from the basic geographic unit to the influence area, the proposed K-TAZ model is improved, the optimal partition problem is more easily solved, and the K-TAZ model is analyzed and derived, In this paper, the lemma and the theorem of the lower bound of the number of the optimal partition of the model are given, and the method of determining the lower bound of the maximum number of partitions is put forward. Based on the homogeneity constraint associated with the number of the partition number in the K-TAZ model, In this paper, the implicit enumeration algorithm for the number of feasible partitions is given as the limit of the number of feasible partitions, and the non-solution scenarios of the K-TAZ model and the theoretical reasons leading to the non-solution of the K-TAZ model are analyzed. In this paper, a heuristic algorithm with two stages of aggregation is designed to reconstruct the solution space so that the K-TAZ model can be solved, and the validity of the proposed algorithm is explained by comparing the heuristic process of the aggregation-type clustering with only one stage and the heuristic process of the proposed algorithm. In this paper, the maximum Kmin domain reduction method is combined with the constraint planning model solving process, and a constraint planning method for solving the K-TAZ model is given, and the method for solving the K-TAZ model is given by combining the feasible partition number of the hidden enumeration domain reduction method with the P median problem model solving process, and the method for solving the P-type model of the K-TAZ model is provided. The solution process and efficiency of two algorithms are compared and analyzed using a simulation numerical example based on Pcb3038 modification. (5) In combination with the actual case of the optimal traffic cell division in the public transport planning of the Suzhou Industrial Park, the application research of the abstract modeling and an example analysis is carried out. The concrete content of this research includes: based on the abstract modeling and calculation method presented in this paper, the data model of the case and the process of establishing the case are given; considering the fixed situation of the number of the partitions, the optimal traffic cell division problem of the case is abstracted into the TAZ-PMP model which minimizes the partition heterogeneity; By using the IE algorithm, the LSLSH algorithm and the LSCG algorithm, the optimal traffic cell division scheme for the case-minimized partition heterogeneity TAZ-PMP model is given, and the efficiency of the algorithm is discussed and analyzed. The optimal traffic cell division problem of the case is abstracted as the K-TAZ model of minimizing the geographical error; the optimal traffic cell division scheme of minimizing the geographic error K-TAZ model is given by using the CP method and the PMP method, and the efficiency of the algorithm solution case is discussed and analyzed.
【学位授予单位】：东北大学
【学位级别】：博士
【学位授予年份】：2014
【分类号】：U491.12;TP301.6
【相似文献】
相关期刊论文 前10条
1 杨波;刘海洲;;基于聚类分析的交通小区划分方法的改进[J];交通与运输(学术版);2007年01期
2 郭峤枫;;浅析交通小区划分问题[J];黑龙江科技信息;2010年28期
3 谭晓雨;;土地利用与交通小区发生吸引量关系研究[J];物流技术;2012年07期
4 谭晓雨;;基于交通小区的道路交通环境负荷发生源分析[J];物流技术;2012年11期
5 马超群;王瑞;王玉萍;严宝杰;陈宽民;;基于区内出行比例的城市交通小区半径计算方法[J];交通运输工程学报;2007年01期
6 李晓丹;杨晓光;陈华杰;;城市道路网络交通小区划分方法研究[J];计算机工程与应用;2009年05期
7 姜培;;基于出行者来源的交通发生与吸引量预测[J];科技广场;2010年03期
8 杜慎旭;;基于新城区土地规划的交通小区出行量预测[J];铁道运输与经济;2012年02期
9 李晓丹;储浩;杨晓光;;城市道路网络交通小区概念解析[J];武汉理工大学学报(交通科学与工程版);2009年05期
10 李雨梦;王晶妍;;聚类分析法确定城市出租车交接班最优地点[J];科协论坛(下半月);2012年10期
相关会议论文 前1条
1 钟章建;黄玮;马万经;姚佼;;面向协调控制的交通小区划分算法设计与实现[A];2008第四届中国智能交通年会论文集[C];2008年
相关博士学位论文 前1条
1 王霖青;交通小区划分问题的整数规划建模与优化算法研究[D];东北大学;2014年
相关硕士学位论文 前6条
1 宋亮;交通小区的理论分析和划分方法研究[D];长安大学;2011年
2 于慧杰;交通小区在交通规划中若干技术问题的研究[D];西安电子科技大学;2008年
3 刘云芳;利用卫星定位系统数据分析交通问题[D];华中师范大学;2012年
4 陈芳;市区对外路网中通道的功能分析及系统配置[D];西南交通大学;2005年
5 刘敏;基于城乡一体化的交通需求分析研究[D];西南交通大学;2008年
6 武明超;基于移动通信网络数据的交通小区划分与OD分析方法研究[D];北京交通大学;2015年
，本文编号：2505495
本文链接：https://www.wllwen.com/kejilunwen/jiaotonggongchenglunwen/2505495.html
