摘要
铁路网车流组织和双向编组站作业分工综合优化问题是铁路运营管理的核心内容之一,其任务是在给定的路网结构、线路和站点能力基础上,制定车流的最优运输组织方案和双向编组站作业分工方案,使车流的运输费用、集结费用和改编费用之和最小。研究者通常将铁路网车流组织问题和双向编组站作业分工综合优化问题进行分别研究,导致获得的研究方案具有局限性。在实际运输过程中,两者之间是相互影响和相互配合的关系,应充分考虑两者之间的内在联系,对它们进行分层或者一体化研究,优化出细化到发到场的直达去向及它们的吸引范围,最大程度的减少双向编组站折角车流、提高编组站利用率、减轻编组站的改编负荷、均衡利用双向系统能力。综上所述,铁路网车流组织和双向编组站作业分工的综合优化具有重要的现实意义。
本文基于车流组织优化理论和现代数学方法,对车流组织和双向编组站作业分工综合优化的有关问题进行了研究。本文的主要研究内容如下：
(1)对比分析了北美地区、欧洲地区和我国的铁路货物运输设备条件、组织现状和典型优化模型,指出了各自的优、缺点及适用范围。
(2)构建了偏好长距离直达去向的列车开行方案优化模型(TFP-c2)。该模型在我国典型货物列车编组计划模型(TFP-c)基础上进行扩展,引入了列车开行频度决策变量,并在目标函数中充分考虑了单位列车运营费用和车流在技术站改编时消耗的人力物力费用。与TFP-c模型的优化方案相比,TFP-c2模型获得的优化方案延长了单位列车的运送距离、减少了车流的中转次数、减轻了编组站的作业负荷、降低了变动设备(机车和车辆)的购置费、加速了货物的送达速度。
(3)构建了考虑列车开行频度和车流树形改编链线性约束的货物列车开行方案优化模型(LTFP)。该模型以欧洲地区的列车编组计划点弧模型为基础,通过引入相同到站车流第一改编站选择决策变量、线性化树形改编链约束和列车频度决策变量,使其适用于我国运输实际。与TFP-c2模型相比,该模型既缩小了问题的求解空问,又便于人工参与优化计算过程。
(4)设计了三种智能优化算法(并行禁忌搜索算法、基于小生境技术的遗传算法、基于邻域搜索和小生境技术的混合遗传算法)对TFP-c、TFP-c2和LTFP模型进行了求解。为验证模型的正确性和评估三种算法的性能,设计了三个不同规模的算例。计算结果表明,在小规模问题上,上述三种优化算法均能求得全局最优解,且求解速度优于商业求解软件中的精确算法(LINGO);当算例规模增大时,智能算法的求解效果均优于精确算法。从求解质量上看,混合遗传算法最好、并行禁忌搜索算法次之、遗传算法最差；从求解效率上看,并行禁忌搜索算法最快、混合遗传次之、遗传算法最差。对比不同固定费用下TFP-c2模型的求解方案发现,列车平均运距随固定费用的增加而增加,达到了预期效果。对比分别在TFP-c2和LTFP模型下的算法求解效果发现,在后者基础上算法的求解效率和求解质量得到极大程度提高。
(5)在满足线路和技术站能力要求下,利用分层优化的方法制定出细化到出发场和到达场的直达方案和车流在双向编组站内的作业分工方案。上层模型为：基于点线网络构建的货物列车编组计划模型(单点模型),下层模型为：双向编组站作业分工优化模型。分层优化中,上层模型优化出的直达去向、车流的改编方案和各直达去向的车流构成作为下层模型的输入数据；下层模型的优化结果作为上层模型的补充和调整依据。
本文设计了三类下层模型,分别为：①到达去向接入系统受限的双向编组站作业分工优化模型；②到发系统不受限的双向编组站作业分工优化模型；③考虑多条折角径路的双向编组站作业分工优化模型。最后,在智能优化算法求解出上层优化模型基础上,利用lingo软件对第②类模型进行了求解,验证了分层优化方法的可行性和模型设计的可行性。
(6)构建了三个不同类型的铁路网车流组织和双向编组站作业分工一体化模型,均以车流的集结费用、改编费用和在双向编组站所在枢纽内联络线的走行费用之和最小为目标。其中前两个模型基于我国车流组织典型模型,分别从仅将路网中标准双向编组站扩展为两个点和将任何形式双向编组站扩展为多点的角度出发,综合考虑双向编组站作业分工问题与路网车流组织问题,构建了一体化模型,后者比前者的应用范围更广泛；第三个模型基于欧洲地区车流组织典型模型,在将路网中标准双向编组站扩展为两点的情形下,增加了符合我国铁路运输实际的具有相同到站的车流需进行树形改编的约束条件,糅合了双向编组站作业分工问题,构建了铁路网车流组织和双向编组站作业分工一体化优化模型。第三个模型较前两个的模型求解空间小,求解速度更快。
最后通过设计算例,验证了上述三个模型的可行性和一致性。计算结果表明,一体化模型得到的优化方案可将分层优化方法中的不可避免折角改编车流转移至其它技术站顺向改编,减少了路网折角改编车流量,减轻了编组站的作业负担和节省了人力物力投资。
Integrated optimizing the train service network design problem and the division of work of two systems of the same marshaling yard problem is one of the core tasks in railway operation. The optimal problem is to obtain railcar flow shipping strategy and railcar flow classification route in marshaling yard in a given network structure, while not exceeding line and yard capacity. The objective is seeks to minimize the sum of shipping, accumulating and classification costs. Researchers usually study the above two problems separated, seldom consider their inner link, lead to optimal result has limitations and adaptability. In the actual transportation process, intense interactions and mutual cooperation exist between the above two problems. When integrate the above two problem into a single optimization model or Hierarchical optimize them, we can set different classification fee for different classification routes, distinguish between different system capabilities and remark connection link miles. It can optimize between which pair of yards provides a direct train service, and the direct train service departure from which system of the start yard and arrive at which system of the end yard? Which car flow routed on this train service? In this method, the railroad can minimize angular wagon flow in bidirectional marshaling stations, improve the marshaling yard utilization, reduce the workload of marshaling yards, and balance the capacity utilization of the two-way system in the same marshaling yards. To sum up, considering the train service network design problem and the division work of two systems of the same marshaling yard problem together has the important practical significance.
Based on the traffic organization theory and modern mathematical methods, this paper make more in-depth study in integrated optimizing the train service network design problem and the division work of two systems of the same marshaling yard problem. The main research content and results are as follows:
(1) Compares and analyzes the railroad transport equipment conditions, organizational status and typical optimization models of North America, European region and China, and pointed out their respective advantages and disadvantages, and the scope of application.
(2) Constructs train service design model preference long distance (TFP-c2). Base on the typical model (TFP-c) of China, the model introduce the train service frequency decision variable, and consider the unit train operation cost and the artificial and equipment cost incurred by the classification work in marshaling yard. Compare the optimal result between model TFP-c and model TFP-c2, it can found that the later can extended transport distance of train services, decrease the number of traffic transfer, release the workload of marshaling station, reduce the fleet size of equipment (locomotive and vehicle), and accelerated the delivery speed of the goods.
(3) Establishes train service design model with train service frequency and linear tree route constraint (LTFP). Base on the typical model of Europe region, the model introduce car flow first classification station selection decision variable, linearization tree chain constraint and train frequency decision variable, which make the model suit our country railway transportation actual. Compare the optimal result between model TFP-c2and model LTFP, it can found that the later not only reduces the solution space, but also convenient for artificial participate in optimization process.
(4) Designs three kinds of intelligent optimization algorithms (parallel tabu search algorithm, genetic algorithm based on niche technique, hybrid genetic algorithm based on the neighborhood search and niche technology) for solving TFP-c, TFP-c2and LTFP models. In order to validate the validity of the model and evaluate the performance of the three algorithms, the paper designs three different scale examples. the calculation result indicates that, in the small scale, the above three kinds of optimization algorithm all can get the global optimal solution with less time consume than accurate algorithm in Commercial solving software (Lingo11); When the example scale increase, Commercial solving software cannot obtain global optimal solution, and the computing time of local optimal solution is longer than that of optimization algorithm. In the view of solution quality, the hybrid genetic algorithm is better than the parallel tabu search algorithm, and the parallel tabu search algorithm is better than genetic algorithm. Comparing the solution result of TFP-c2model with different train dispatching cost, we can find that average train running distance is growing with the increasing cost. Compare the solution effect of intelligent optimization algorithms in model LTFP to in model TFP-c2, it can be found that the former model can greatly reduce solution space.
(5) Proposes a hierarchical optimization method to determine between which pair of yards provides a direct train service, and the direct train service departure from which system of the start yard and arrive at which system of the end yard? The classification routes of cars between two systems of a marshaling yard. The up model is train service design model in which classification yard represent as a single node, rail tracks represent as lines, and the lower model is to determine classification routes of cars in a marshaling yard with two systems. In the optimal process, the optimal result of the up model can be used as input data to the lower model, the solution result of the lower can be as the supplement and adjustment of the upper model.
This paper designs three types of the lower model, respectively:①In the arriving yard of a train service, the service is limited to access only one system of a marshalling station with two system in advance, in this condition, constructed a model for operation division of work of two systems in a marshaling yard.②In the arriving or departing yard of a train service, trains are not limited to out or access only one system in advance, in this condition, developed a model for operation division of work of two systems in a marshaling yard.③take multiple angular path of car flow classification route in bidirectional marshalling station, developed a model for operation division of work of two systems in a marshaling yard. Finally, using the intelligent optimization algorithms to solve the upper optimization model, and using the commercial software (lingo) solve to the second lower model, verify the hierarchical optimization method is feasible and the model design feasibility.
(6) Constructs three different types of integrate models for a combined train service design and division of work of two system in a marshalling yard problem aiming at minimizing train service accumulated delay cost, classification cost and shipping cost in connection path in hub yard. The first two models are based on typical model of China, the former extend standard marshalling station with two systems to two nodes, the later extend Arbitrary structure marshalling station with two systems to several nodes, then consider the division of work of two system in a marshalling yard problem into train service design problem, constructed two type integrated models. The later has more extensive application scope than the former. The last one model is based on typical model of Europe, adding linear constraint of flow tree shipping chain which make the model suit the actual situation of China Railway system, simultaneous address the division of work of two system in a marshalling yard problem and train service design problem in railroad, developed another integrated model. The last one can reduce solution scope than the other two.
Finally an example was presented to validate the feasibility and the consistency of the above three models. The results show that integration model can change car flow shipping strategy and avoid inevitable angle classification car flows obtaining from hierarchical optimization method, reduce the total number of angle classification cars, release the workload of marshaling yards and saving human resources investment.
引文
[1]王会宗,纪涛.对我国铁路运输业垄断程度的重新认识[J].天府新论,2008(5)：58-62.
[2]李建平.铁路货物运输的现状及发展[J].中国铁路,2002(8)：46-47.
[3]关伟.传统铁路货运与现代物流业[J].铁路采购与物流,2011,06(8)：54-55.
[4]杨浩.铁路运输组织学[M].北京：中国铁道出版社,2006.
[5]宋晓东,陈治亚.铁路双向编组站的折角车流研究[J].铁道运营技术,2009,15(2)：39-42.
[6]曹魁久,孔庆钤.货物列车编组计划[M]：北京：中国铁道出版社,1992.
[7]杨浩.铁路运输组织学[M].北京：中国铁道出版社,2006.
[8]ASSAD, ARJANG A.. Models for rail transportation[J]. Transportation Research Part a: General,1980,14(3):205-220.
[9]ASSAD, ARJANG A.. Modelling of rail networks:Toward a routing/makeup model[J]. Transportation Research Part B:Methodological,1980,14(1/2):101-114.
[10]BODIN L D, GOLDEN B L, SCHUSTER A D, et al. A model for the blocking of trains[J]. Transportation Research Part B:Methodological,1980,14(1):115-120.
[11]ASSAD A A. Analysis of rail classification policies[J]. Infor,1983,21(4):293-314.
[12]VAN DYKE C D. The automated blocking model:A practical approach to freight railroad blocking plan development[J],1986.
[13]VAN DYKE C D. Dynamic management of railroad blocking plans[J],1988.
[14]NEWTON H N. Network design under budget constraints with application to the railroad blocking problem[J]:AIR FORCE INST OF TECH WRIGHT-PATTERSON AFB OH, 1997.
[15]NEWTON H N, BARNHART C, VANCE P H. Constructing railroad blocking plans to minimize handling costs[J]. Transportation Science,1998,32(4):330-345.
[16]BARNHART C, JIN H, VANCE P H. Railroad blocking:A network design application[J]. Operations Research,2000:603-614.
[17]AHUJA R K, JHA K C, LIU J. Solving real-life railroad blocking problems[J]. Interfaces, 2007,37(5):404-419.
[18]AHUJA R K, JHA K C, ORLIN J B, et al. Very large-scale neighborhood search for the quadratic assignment problem[J]. Informs Journal on Computing,2007,19(4):646-657.
[19]AHUJA R K, CUNHA C B, SAHIN G. Network models in railroad planning and scheduling[J]. Tutorials in Operations Research,2005,1:54-101.
[20]MASOUD Y, FOROUGHI A, NADJARIB. Solving railroad blocking problem using ant colony optimization algorithm[J]. Applied Mathematical Modelling,2011,35(12): 5579-5591.
[21]LIU J, AHUJA R K. Optimal network configuration and capacity expansion of railroads[J]. Journal of the Operational Research Society,2007,59(7):911-920.
[22]THOMET M A. A user-oriented freight railroad operating policy[J]. Systems, Man and Cybernetics, IEEE Transactions on,1971,1(4):349-356.
[23]CRAINIC T, FERLAND J A, ROUSSEAU J M. A tactical planning model for rail freight transportation[J]. Transportation Science,1984,18(2):165-184.
[24]KEATON M H. Designing optimal railroad operating plans:Lagrangian relaxation and heuristic approaches[J]. Transportation Research Part B:Methodological,1989,23(6): 415-431.
[25]KEATON M H. Service-cost tradeoffs for carload freight traffic in the US rail industry[J]. Transportation Research Part a:General,1991,25(6):363-374.
[26]KEATON M H. Designing railroad operating plans:A dual adjustment method for implementing Lagrangian relaxation[J]. Transportation Science,1992,26(4):263-279.
[27]HAGHANI A E. Formulation and solution of a combined train routing and makeup, and empty car distribution model[J]. Transportation Research Part B:Methodological,1989, 23(6):433-452.
[28]MARTINELLI D R, TENG H. Neural network approach for solving the train formation problem[J]. Transportation Research Record,1996,4(1):33-49.
[29]MARIN A, SALMERON J. Tactical design of rail freight networks. Part I:Exact and heuristic methods[J]. European Journal of Operational Research,1996,90(1):26-44.
[30]MARIN A, SALMERON J. Tactical design of rail freight networks. Part II:Local search methods with statistical analysis[J]. European Journal of Operational Research,1996,94(1): 43-53.
[31]FUKASAWA R, POGGI D V, PORTO O, et al. Solving the freight car flow problem to optimality[J]. Electronic Notes in Theoretical Computer Science,2002,66(6):42-52.
[32]JEONG S J, LEE C G, BOOKBINDER J H. The European freight railway system as a hub-and-spoke network[J]. Transportation Research Part a:Policy and Practic,2007,41(6): 523-536.
[33]JHA K C, AHUJA R K, SAHIN G. New approaches for solving the block-to-train assignment problem[J]. Networks,2008,51(1):48-62.
[34]ANDERSEN J, CRAINIC T G, CHRISTIANSEN M. Service network design with asset management:Formulations and comparative analyses[J]. Transportation Research Part C: Emerging Technolog,2009,17(2):197-207.
[35]ANDERSEN J, CRAINIC T G, CHRISTIANSEN M. Service network design with management and coordination of multiple fleets [J]. European Journal of Operational Research,2009,193(2):377-389.
[36]LULLI G, PIETROPAOLI U, RICCIARDIN. Service network design for freight railway transportation:the Italian case[J]. Journal of Operational Research Society,2011,62: 2017-2119.
[37]YAGHINI, MASOUD. Multicommodity network design problem in rail freight transportation planning[J]. Procedia-Social and Behavioral Sciences,2012,43(0):728-739.
[38]YAGHINI M, KARIMI M, RAHBAR M, et al. A hybrid simulated annealing and simplex method for Fixed-Cost capacitated multicommodity network design[J]. International Journal of Applied Metaheuristic COM,2012,2(4):13-28.
[39]YAGHINI M, MOMENI M, SARMADI M. A simplex-based simulated annealing algorithm for node-arc capacitated multicommodity network design[J]. Applied Soft Computing,2012, 9(12):2997-3003.
[40]YAGHINI M, MOMENI M, SARMADI M. An improved local branching approach for train formation planning[J]. Applied Mathematical Modelling,2012, in press.
[41]YAGHINI M, AKHAVAN M R. A simulated annealing algorithm for unsplittable capacitated network design[J]. International Journal of Industrial Engineering,2012,23(2): 91-100.
[42]朱松年.运筹学在编制货物直达运输方案中的应用[J].铁道运营技术,1995(2)：76-77,89.
[43]朱松年,曹家明,赵强,等.车流组织综合优化[J].铁道学报,1993(3)：59-69.
[44]曹家明.铁路网上车流组织的综合优化[D].成都：西南交通大学,1992.
[45]曹家明.大规模列车编组计划问题的优化算法[C]//中国科学技术协会首届青年学术年会：中国科学技术协会,1992.
[46]曹家明,朱松年.优化列车编组计划的0-1规划法[J].铁道学报,1992(4)：49-57.
[47]曹家明,朱松年.铁路网上列车编组计划的综合编制[J].西南交通大学学报,1993(5)：93-98.
[48]曹家明,朱松年.铁路网上技术直达列车编组计划优化的二次0-1规划法[J].铁道学报,1993(2)：62-70.
[49]林柏梁,朱松年,赵强.技术直达列车编组计划的阶跃函数模型及同构变换[J].西南交通大学学报,1994(1)：91-96.
[50]林柏梁,朱松年,史德耀,等.装车地直达列车编组计划的优化模型[J].中国铁道科学,1995(2)：108-114.
[51]林柏梁,朱松年,陈竹生,等.路网上车流径路优化的0-1规划模型及其合理径路集生成算法[J].铁道学报,1997(1)：9-14.
[52]林柏梁,朱松年.路网上车流径路与列车编组计划的整体优化[J].铁道学报,1996(1)：1-7.
[53]林柏梁,朱松年.带有改编能力限制的编组计划优化模型及算法[J].西南交通大学学报,1994(5)：488-493.
[54]林柏梁,朱松年.优化编组计划的非线性0—1规划模型及模拟退火算法[J].铁道学报,1994(2)：61-66.
[55]林柏梁,徐忠义,黄民,等.路网发展规划模型[J].铁道学报,2002,24(2)：1-6.
[56]林柏梁,田亚明,王志美.基于最远站法则的列车编组计划优化双层规划模型[J].中国铁道科学,2011,32(5)：108-113.
[57]林柏梁,乔国会.基于线路能力约束下的铁路空车调配迭代算法[J].中国铁道科学,2008,29(1)：93-96.
[58]林柏梁,陈竹生.路网编组站间重载列车去向优化分布模型[J].铁道学报,1998(2)：20-24.
[59]林柏梁.机车长交路条件下的技术站列车编组计划无调作业参数模型[J].铁道学报,1999(6)：6-11.
[60]林柏梁.车流运行径路与列车编组计划的整体优化模型及模拟退火算法[D].成都：西南交通大学,1994.
[61]林柏梁.直达与区段列车编组计划及车流径路的整体优化方法[J].铁道学报,1996,18(5)：1-8.
[62]史峰,莫辉辉,黎新华,等.区段管内车流组织优化方法[J].交通运输工程学报,2003,3(2)：84-87.
[63]史峰,李致中.列车编组计划网络优化方法[J].铁道学报,1994,16(2)：74-79.
[64]史峰.铁路车流组织优化[D].北京：北方交通大学北京交通大学,1998.
[65]彭其渊,王培.网络列车运行图的数学模型及算法研究[J].铁道学报,2001(1)：1-6.
[66]查伟雄,秦作睿.路网上列车编组计划的有利去向模型及其网络流算法[J].铁道学报,1998(2)：12-18.
[67]查伟雄,熊桂林,张广洲.列车编组计划模型及其计算策略[J].长沙铁道学院学报,2001,19(4)：27-30.
[68]查伟雄.列车编组计划决策支持系统的研究[D].北京：北方交通大学北京交通大学,1997.
[69]许红,马建军,龙昭,等.技术站单组列车编组方案模型与计算方法的研究[J].铁道学报,2006(3)：1-5.
[70]王保华,何世伟.铁路车流改编方案随机优化模型及其算法[J].中国铁道科学,2009(05)：104-108.
[71]李宵寅.基于不确定参数的列车编组计划与车流径路综合优化研究[D].成都：西南交通大学,2011.
[72]曹学明,林柏梁,严贺祥.铁路直达运输条件下煤炭运输与库存一体化模型[J].北京交通大学学报(自然科学版),2006,30(6)：27-31.
[73]纪丽君.基于物流成本的装车地车流组织优化方法研究[D].北京：北京交通大学,2008.
[74]强丽霞.基于端点换重的装车地直达车流组织优化模型与算法研究[J].铁道学报,2009,31(6)：91-96.
[75]范振平,林柏梁,牛会想.装车地始发直达开行方案研究[J].物流技术,2007,26(6)：77-80.
[76]范振平,乐逸祥,范登科,等.基于蚁群算法的装车地直达重载运输优化模型求解算法研究[J].物流技术,2007,26(8)：110-114,13.
[77]YUE Yi-xiang, ZHOU Lei-shan, YUE Qun-xing, et al. Multi-route railroad blocking problem by improved model and ant colony algorithm in real world[J]. Computers & Amp; Industrial Engineering,2011,60(1):34-42.
[78]LIN B L, WANG Z M. Optimizing the freight train connection service network of a large-scale rail system[J]. Transportation Research Part B:Methodological,2012,46(5): 649-667.
[79]黎浩东,何世伟,王保华,等.铁路车流改编方案与技术站布局的综合优化[J].铁道学报,2011(02)：8-12.
[80]黎浩东,何世伟,宋瑞,等.列车编组计划和技术站布局的综合优化[J].北京交通大学学报,2010,34(6)：30-34,39.
[81]田亚明,林柏梁,王志美.考虑始发直达比重的编组站改编能力优化模型[J].交通运输系统工程与信息,2011(03)：65-70.
[82]曹学明.加速货车循环相关运输组织优化模型与方法研究[D].北京：北京交通大学,2007.
[83]曹学明,王喜富,林柏梁.重空车流径路及多车种空车调配协同优化模型[J].中国铁道科学,2009,30(6)：114-118.
[84]田亚明,林柏梁,纪丽君.基于多商品流和虚拟弧的铁路车流分配点-弧、弧-路模型研究[J].铁道学报,2011(04)：7-12.
[85]纪丽君,林柏梁,乔国会,等.基于多商品流模型的铁路网车流分配和径路优化模型[J].中国铁道科学,2011,32(3)：107-110.
[86]李夏苗,卢红岩.快运货物列车编组方案的优化[J].中国铁道科学,2004(02)：118-121.
[87]梁栋,林柏梁.技术站分组列车编组计划的优化理论及模型研究[J].铁道学报,2006(03)：1-5.
[88]赵鹏,张进川,唐宝刚.基于组合列车的重载铁路装车区车流组织优化模型研究[J].中国铁道科学,2010,31(6)：116-121.
[89]王志美,林柏梁,刘希元.牵引定数不统一的港口后方腹地车流组织优化[J].北京交通大学学报,2011(03)：62-67.
[90]左武,符卓,肖天国.货物列车编组计划的自适应遗传算法研究[J].石家庄铁道大学学报,2011,24(1)：84-90.
[91]左武.铁路分组货物列车编组计划的遗传算法研究[D].长沙：中南大学,2009.
[92]陈崇双,王慈光,杨运贵,等.不确定条件下分组列车在编成站的集结参数研究[J].铁道学报,2011(05)：1-7.
[93]陈崇双,王慈光,杨运贵,等.不确定条件下开行固定车组重量分组列车适用条件研究[J].铁道学报,2011(12)：1-8.
[94]张全寿.编组站站调决策支持系统的研究[J].北方交通大学学报,1989(04)：64-72.
[95]牛惠民,胡安洲.铁路枢纽车流组织的非线性0-1规划模型及算法[J].铁道学报,2001,23(3)：8-12.
[96]牛惠民.双向编组站系统作业分工的优化模型及算法[J].中国科学：E辑,1998,28(004)：378-384.
[97]牛惠民.铁路枢纽重空车流组织的协调优化模型及遗传算法[J].铁道学报,2001(4)：12-17.
[98]GULBRODSEN O. Optimal planning of marshalling yard by operation research[J],1963: 226-233.
[99]LENTINK R M, FIOOLE P J, KROON L G, et al. Applying operations research techniques to planning of train shunting[J]. Planning in Intelligent Systems,2003:415-436.
[100]HEIN O. A two-stage queue model for a marshalling yard[J]. Rail International,1972,3(4): 249-250.
[101]PETERSEN E. Railyard modeling:Part I. Prediction of put-through time[J]. Transportation Science,1977,11(1):37-49.
[102]PETERSEN E. Railyard modeling:Part II. The effect of yard facilities on congestion[J]. Transportation Science,1977,11(1):50-59.
[103]TURNQUIST M A, DASKIN M S. Queuing models of classification and connection delay in railyards[J]. Transportation Science,1982,16(2):207-230.
[104]MARTLAND C D. PMAKE analysis:predicting rail yard time distributions using probabilistic train connection standards[J]. Transportation Science,1982,16(4):476-506.
[105]曹家明,范征,毛节铭.编组站作业优化决策支持系统——解体子系统[J].铁道学报,1993(4)：66-73.
[106]刘军.铁路编组站作业计划滚动式智能化编制方法的研究[D].北京：北方交通大学北京交通大学,1993.
[107]何世伟,宋瑞,朱松年.编组站阶段计划解编作业优化模型及算法[J].铁道学报,1997(3)：2-9.
[108]何世伟,宋瑞,朱松年.铁路编组站阶段计划编制的模型及其算法研究[J].系统工程理论与实践,1997(2)：89-95.
[109]何世伟,宋瑞,朱松年.铁路编组站日班计划网络流规划模型及算法[J].铁道学报,1995(4)：8-15.
[110]何世伟,宋瑞,胡安洲.编组站阶段计划刚性与柔性优化的协调研究[J].铁道学报,1999(4)：1-8.
[111]何世伟.铁路枢纽作业计划的优化编制--编组站作业计划与枢纽日班计划的模型及算法研究[D].成都：西南交通大学,1996.
[112]王慈光.用表上作业法求解编组站配流问题的研究[J].铁道学报,2002(04)：1-5.
[113]王慈光.编组站动态配流模型与算法研究[J].铁道学报,2004(01)：1-6.
[114]王慈光.编组站静态配流网络模型[J].交通运输工程与信息学报,2003(02)：67-71.
[115]邢文训,谢金星.现代优化计算方法[M].北京：清华大学出版社,2005.
[116]GLOVER F. Tabu search-part II[J]. ORSA Journal on Computing,1990,2(1):4-32.
[117]GLOVER F. Tabu search-part I[J]. ORSA Journal on Computing,1989,1(3):190-206.
[118]GLOVER F. Tabu search:A tutorial[J]. Interfaces,1990,20(4):74-94.
[119]GLOVER F. Future paths for integer programming and links to artificial intelligence[J]. Computers & Operations Research,1986,13(5):533-549.
[120]BADEAU, PHILIPPE. A parallel tabu search heuristic for the vehicle routing problem with time Windows[J]. Transportation Research Part C:Emerging Technolog,1997,5(2): 109-122.
[121]JIN, JIANYONG. A parallel multi-neighborhood cooperative tabu search for capacitated vehicle routing problems[J]. European Journal of Operational Research,2012,222(3): 441-451.
[122]ATTANASIO, ANDREA. Parallel Tabu search heuristics for the dynamic multi-vehicle dial-a-rideproblem[J]. Parallel Computing,2004,30(3):377-387.
[123]CARICATO, PIERPAOLO. Parallel tabu search for a pickup and delivery problem under track contention[J]. Parallel Computing,2003,29(5):631-639.
[124]GENDREAU, MICHEL. A dynamic model and parallel tabu search heuristic for real-time ambulance relocation[J]. Parallel Computing,2001,27(12):1641-1653.
[125]FIECHTER, C.-N.. A parallel tabu search algorithm for large traveling salesman problems[J]. Discrete Applied Mathematics,1994,51(3):243-267.
[126]AUGUGLIARO, A.. An evolutionary parallel TABU search approach for distribution systems reinforcement planning[J]. Advanced Engineering Informatics,2002,16(3): 205-215.
[127]JAMES, TABITHA. A cooperative parallel tabu search algorithm for the quadratic assignment problem[J]. European Journal of Operational Research,2009,195(3):810-826.
[128]PORTO, STELLA C.S.. Performance evaluation of a parallel tabu search task scheduling algorithm[J]. Parallel Computing,2000,26(1):73-90.
[129]BORTFELDT, A. A parallel tabu search algorithm for solving the container loading problem[J]. Parallel Computing,2003,29(5):641-662.
[130]CRAINIC T G, TOULOUSE M, GENDREAU M. Toward a taxonomy of parallel tabu search heuristics[J]. Informs Journal on Computing,1997,9:61-72.
[131]王小平,曹立明.遗传算法—理论、应用于软件实现[M].西安：西安交通大学出版社,2002.
[132]付强,王刚,王明宇,等.基于小生境遗传算法的制导雷达误差估计[J].空军工程大学学报(自然科学版),2011,12(6)：50-53.
[133]李红婵,朱颢东.基于小生境遗传算法的排课问题研究[J].计算机工程,2011,37(16)：194-196.
[134]田琳琳,赵宁,钟伟,等.基于小生境遗传算法的风电场布局优化[J].南京航空航天大学学报,2011,43(5)：650-654.
[135]李隽颖,楼晓俊.基于小生境遗传算法的分类优化方法[J].计算机应用研究,2012,29(5)：1787-1790.
[136]ERICKSON M, MAYER A, HORN J. Multi-objective optimal design of groundwater remediation systems:application of the niched Pareto genetic algorithm (NPGA)[J]. Advances in Water Resources,2002,25(1):51-65.
[137]ABIDO M A. A niched Pareto genetic algorithm for multiobjective environmental/economic dispatch[J]. International Journal of Electrical Power & Amp; EN,2003,25(2):97-105.
[138]GRANDINETTIL, GUERRERO F, LEPERA G, et al. A niched genetic algorithm to solve a pollutant emission reduction problem in the manufacturing industry:A case study[J]. Computers & Amp; Operations Research,2007,34(7):2191-2214.
[139]ROY M, SHAFAIC. A MEM electric field sensor optimization by Multi-Objective niched pareto genetic algorithm[J]. Procedia Engineering,2011,25(0):1609-1612.
[140]崔挺,孙元章,徐箭,等.基于改进小生境遗传算法的电力系统无功优化[J].中国电机工程学报,2011,31(19)：43-50.
[141]周北岳,邓斌,郭观七.基于小生境技术的改进遗传算法研究[J].机械强度,2002(01)：13-16+103.
[142]徐金梧,刘纪文.基于小生境技术的遗传算法[J].模式识别与人工智能,1999(01)：104-108.
[143]赵苗,齐丽丽,李建晶.基于小生境技术的遗传优化算法改进[J].电脑与电信,2010(01)：39-40.
[144]张振安.基于小生境技术改进遗传算法在供电网规划中的应用[D].南京：东南大学,2005.
[145]高岚,胡友为,潘峰,等.基于小生境遗传算法的SAR图像分割[C]//可持续发展的中国交——2005全国博士生学术论坛(交通运输工程学科)论文集(下册).中国北京,2005：1033-1034.
[146]石明奎,周强,王小平.基于小生境遗传算法的排课系统的设计与研究[J].洛阳理工学院学报(自然科学版),2010(02)：52-55.
[147]刘国炜,常新功.小生境技术在遗传规划中的应用[J].计算机系统应用,2011(02)：49-52+156.
[148]喻寿益,郭观七.一种改善遗传算法全局搜索性能的小生境技术[J].信息与控制,2001(06)：526-530+54.
[149]乔佩利,郑林,马丽丽.一种小生境遗传算法研究[J].哈尔滨理工大学学报,2011(01)：90-93.
[150]王亚子,贾利新.遗传算法的小生境技术改进[J].河南教育学院学报(自然科学版),2008(01)：28-29.
[151]田青,周里群,熊根发,等.自适应小生境混合遗传算法在车间调度中的应用研究[J].机电产品开发与创新,2010(04)：26-28.
[152]赵赫,杜端甫.TSP的邻域搜索算法的分析和改进[J].中国管理科学,1997,5(1)：35-39.
[153]王凌,郑大钟.邻域搜索算法的统一结构和混合优化策略[J].清华大学学报(自然科学版),2000(09)：125-128.
[154]欧锦文,施保昌.平行机排序邻域搜索算法设计[J].计算机工程与应用,2003(18)103-104+11.
[155]陈昊.求解NP难问题的邻域搜索方法[J].湖北大学成人教育学院学报,2005(01)：80-81.
[156]陈云花,李静.基于邻域搜索的快速分形图像编码算法[J].中国高新技术企业,2009 (21)：20-21.
[157]张亚玲.一种邻域搜索算法在差异工件单机批调度问题中的应用研究[D].合肥：中国科学技术大学,2010.
[158]王晶,姚辉,王艳亮.有限等待流水车间调度邻域搜索算法[J].工业工程与管理,2010(03)：55-59.
[159]张勇军,任震,钟红梅,等.实时无功优化调度中的邻域搜索改进遗传算法[J].电网技术,2003,27(1)：22-25.
[160]顾雷,席裕庚.一种并行多目标遗传邻域搜索算法[J].控制工程,2009(06)：738-742.
[161]王振禄,蔡慧林.资源受限项目调度中基于邻域搜索的混合遗传算法[J].兰州交通大学学报,2005(03)：12-14.
[162]陆子强,郭国雄,蒋金山.基于邻域搜索的混合遗传算法及其在对称TSP中的应用[J].计算机工程与应用,2005(07)：79-81+96.