公铁高速客运方式分担率模型研究
|
摘要
近几年来,高速铁路在我国迅速发展起来。高速铁路的迅猛发展对传统的客运方式造成很大的冲击。高速班车是在高速列车出现之前,地面主要的快速客运方式之一,由于其班次多、速度快,占有着相当的客运市场份额。高速列车引入之后,有一小部分原高速班车的乘客转移到高速列车,两者在不同的距离及服务水平下存在着一定的竞争关系。目前大部分研究着眼于公路与铁路的协调关系,对高速班车、高速列车之间的竞争关系的研究比较少。
本研究着眼于陆路交通的高速客运体系,对公铁高速班车、高速列车两种高速客运方式进行了分担率研究。
首先,从宏观角度分析了为保证模型的精度对RP+SP调查得到的原始数据进行校正,得到与实际符合的工程数据。然后本研究从出行者特性、交通需求特性、选择枝特性三个方面分析了影响出行得交通方式选择的主要因素,并从宏观的角度分析了区域范围内的经济、地域对于出行者交通方式选择行为的影响。基于因素分析建立公铁高速客运方式分担率模型,提出运用指示变量法、定性排序变量得分法、对比标度权重系数法将定性变量量化。
紧接着,本研究基于共同OD点对建立了公铁高速列车、高速班车两种快速客运方式分担率模型。建立将广义费差——时间差、费用差之和作为效用函数中可以被观测到的部分的分担率模型,提出了基于Excel宏(VBA)编程的最大似然函数求解方法求解模型参数。
本研究以浙江省甬台温走廊上的高速班车、高速列车两种高速客运方式为依托,分别采用基于因素分析的分担率模型以及基于共同OD点对的logit模型。基于因素分析的分担率模型研究确定到站时间、月收入、职业类型、票价为影响高速列车选择的主要因素,而出行频率对于分担率基本上没有影响,并通过算例对其进行了验证,该模型同时验证了采用指标变量、定序变量定量化及对比权重法等将定性指标量化方法的正确性；基于共同OD点对的模型确定了以高速列车、高速班车在各个距离范围内的分担率,并确定以150km作为高速班车及高速列车适应出行距离的分界点。这与国外研究中高速列车的适应距离为200km-500km、高速班车的适应距离为200km以下的结论基本一致。并且,适应出行距离的结论验证了基于因素分析的分担率模型的算例的正确性。
最后,在实例应用及调查分析的基础上,提出了对高速班车客运发展的几点建议。
High-speed railway (HSR) in China has developed rapidly in recent years, which causes great impact on traditional passenger transport modes. High-speed bus (HSB) is one of the land rapid transport modes, having a great market share because of its speed, convenience and frequent shifts.
After the introduction of HSR, part of the original HSB passengers transferred to the HSR, but after the transfer, the passenger flow is basically stable within certain distance. Currently, most research focused on the coordination between road and rail, and HSR, the study of competition between HSR and HSB is relatively not that much. This study focuses on the rapid passenger transport system, and builds a diversion model between two fast transport modes-HSB&HSR. The following parts explain the details:
First, this study analyzes the factors that influence the choice behavior of the travelers from three parts i.e. travelers'characteristics, travel demand and characteristics of alternatives. The economy and the location within the region are also analyzed from a macro point of view. And based on this analysis, a factor analysis based sharing model is established. As there are many qualitative variables in the sharing model, three methods to quantify them are proposed: Indicators (dummy variables), scores of ordered qualitative, and weight coefficient.
Flowingly, this paper discusses the building method of sharing model of HSR and HSB based on mutual OD pairs. First, it analyzes the necessity of data validation and expansion. And it provides the method to get engineering data. Then with time and cost differences as the visible parts of utility function, an Excel VBA based program is developed to estimate the parameters of Mutual OD pairs based model.
This paper applies both models to Yongtaiwen corridor. The result of the application of model based on factor analysis indicates that the access time, monthly income, career and ticket fare as the main factors that affect the sharing rate of HSR, which also proves travel frequency has no effect on the sharing rate as well as the validity of the methods to quantize the qualitative variables; While the application of mutual OD pairs based model identifies the suitable travel distance scope of HSR and HSB. It concludes that 150km as division, i.e. if the distance is less than 150km, HSB proves a good passenger transport mode, while of the distance more than 150km, HSR is a good choice for the passengers. This conclusion is in line with the research of experts, which uses 200km as a division. And moreover the result of mutual OD pairs model test the result of calculation example of the factor analysis based model.
Finally, based on the application of both models and the research analysis, some suggestions are made to encourage the development of HSB.
本研究着眼于陆路交通的高速客运体系,对公铁高速班车、高速列车两种高速客运方式进行了分担率研究。
首先,从宏观角度分析了为保证模型的精度对RP+SP调查得到的原始数据进行校正,得到与实际符合的工程数据。然后本研究从出行者特性、交通需求特性、选择枝特性三个方面分析了影响出行得交通方式选择的主要因素,并从宏观的角度分析了区域范围内的经济、地域对于出行者交通方式选择行为的影响。基于因素分析建立公铁高速客运方式分担率模型,提出运用指示变量法、定性排序变量得分法、对比标度权重系数法将定性变量量化。
紧接着,本研究基于共同OD点对建立了公铁高速列车、高速班车两种快速客运方式分担率模型。建立将广义费差——时间差、费用差之和作为效用函数中可以被观测到的部分的分担率模型,提出了基于Excel宏(VBA)编程的最大似然函数求解方法求解模型参数。
本研究以浙江省甬台温走廊上的高速班车、高速列车两种高速客运方式为依托,分别采用基于因素分析的分担率模型以及基于共同OD点对的logit模型。基于因素分析的分担率模型研究确定到站时间、月收入、职业类型、票价为影响高速列车选择的主要因素,而出行频率对于分担率基本上没有影响,并通过算例对其进行了验证,该模型同时验证了采用指标变量、定序变量定量化及对比权重法等将定性指标量化方法的正确性；基于共同OD点对的模型确定了以高速列车、高速班车在各个距离范围内的分担率,并确定以150km作为高速班车及高速列车适应出行距离的分界点。这与国外研究中高速列车的适应距离为200km-500km、高速班车的适应距离为200km以下的结论基本一致。并且,适应出行距离的结论验证了基于因素分析的分担率模型的算例的正确性。
最后,在实例应用及调查分析的基础上,提出了对高速班车客运发展的几点建议。
High-speed railway (HSR) in China has developed rapidly in recent years, which causes great impact on traditional passenger transport modes. High-speed bus (HSB) is one of the land rapid transport modes, having a great market share because of its speed, convenience and frequent shifts.
After the introduction of HSR, part of the original HSB passengers transferred to the HSR, but after the transfer, the passenger flow is basically stable within certain distance. Currently, most research focused on the coordination between road and rail, and HSR, the study of competition between HSR and HSB is relatively not that much. This study focuses on the rapid passenger transport system, and builds a diversion model between two fast transport modes-HSB&HSR. The following parts explain the details:
First, this study analyzes the factors that influence the choice behavior of the travelers from three parts i.e. travelers'characteristics, travel demand and characteristics of alternatives. The economy and the location within the region are also analyzed from a macro point of view. And based on this analysis, a factor analysis based sharing model is established. As there are many qualitative variables in the sharing model, three methods to quantify them are proposed: Indicators (dummy variables), scores of ordered qualitative, and weight coefficient.
Flowingly, this paper discusses the building method of sharing model of HSR and HSB based on mutual OD pairs. First, it analyzes the necessity of data validation and expansion. And it provides the method to get engineering data. Then with time and cost differences as the visible parts of utility function, an Excel VBA based program is developed to estimate the parameters of Mutual OD pairs based model.
This paper applies both models to Yongtaiwen corridor. The result of the application of model based on factor analysis indicates that the access time, monthly income, career and ticket fare as the main factors that affect the sharing rate of HSR, which also proves travel frequency has no effect on the sharing rate as well as the validity of the methods to quantize the qualitative variables; While the application of mutual OD pairs based model identifies the suitable travel distance scope of HSR and HSB. It concludes that 150km as division, i.e. if the distance is less than 150km, HSB proves a good passenger transport mode, while of the distance more than 150km, HSR is a good choice for the passengers. This conclusion is in line with the research of experts, which uses 200km as a division. And moreover the result of mutual OD pairs model test the result of calculation example of the factor analysis based model.
Finally, based on the application of both models and the research analysis, some suggestions are made to encourage the development of HSB.
引文
[1]Yung-Hsiang Cheng. High-speed rail in Taiwan New experience and issues for future development [J].Transport Policy,2010, V (17):51-63
[2]高速时代风起云涌公路客运何去何从.交通世界,2009年第12期.
[3]中长期铁路网规划(2008年调整)。
[4]Benedikt Mandel, Marc Gaudry, Werner Rothengatter. (1994) linear or nonlinear utility functions in logit models? The impact on German high-speed rail demand forecasts [J].Transpn.Res.28B (2),91-101
[5]PETER NIJKAMP, AURA REGGIANI, TOMMASO TRITAPEPE:MODELLING INTER-URBAN TRANSPORT FLOWS IN ITALY:A COMPARISON BETWEEN NEURAL NETWORK ANALYSIS AND LOGIT ANALYSIS [J]. Trans Res.-C, Vol.4, No.6, pp.323-338,1996
[6]Agostino Nuzzolo, Umberto Crisalli, Francesca Gangemi. A behavioural choice model for the evaluation of railway supply and pricing policies[J]. Transportation Research Part A,2000, V(34):395-404
[7]Moshe Ben-Akiva, Takayuki Morikawa. Comparing ridership attraction of rail and bus[J]. Transport policy,2002, V9:107-116
[8]Enjian Yao T. Morikawa, S. Kurauchi, T. Tokida. A study on nested logit mode choice model for intercity high-speed rail system with combined RP/SP data [J]. Proceedings of the Conference on Traffic and Transportation Studies, ICTTS, p 612-619,2002
[9]Gonzalez-Savignat, Mar (2004)'Will the high-speed train compete against the private vehicle?',Transport Reviews,24:3,293—316
[10]Chang, Iljoon and Chang, Gang-len(2004)'A network-based model for estimating the market share of a new high-speed rail system', Transportation Planning and Technology, 27:2,67—90
[11]Yonghwa Park, Hun-Koo Ha. Analysis of the impact of high-speed railroad service on air transport demand[J].Transportation Research Part E 42 (2006) 95-104
[12]石定宇.运输方式选择行为模型的参数估计方法[J].西南交通大学学报,1991,4：1-7
[13]石定宇,高旭敏.运输通道中公路与铁路旅客选择行为模型研究[J].铁道学报,1992,14(2)：58-63
[14]杨兆升,黎春兰.客运方式选择方法研究[J].中国公路学报,1997,10(2)
[15]黄承锋,杨秀苔,李毅.运输通道市场结构变动的静态博弈分析——以成渝通道客运为例[J].重庆交通学院学报,2000,19(4)：105-109
[16]胡文友,李旭宏.公路运输通道内的车辆出行路径选择模型及应用研究[J].公路交通科技,2003,20(1)：125-128
[17]马波涛,张于心,赵翠霞.运用Logit模型对高速客流分担率的估计[J].北方交通大学学报,2003,27(2)：66-69
[18]孙海林,胡兆杰.秦沈客运专线对京沈客流分担率的影响分析[J].铁道运输与经济研究与建议,2004,26(3)：69-71
[19]谢如鹤,邱祝强,李庆云,王荣华.Logit模型在广深铁路客流分担率估算中的应用[J].中国铁道科学,2006,27(3)：111-114
[20]张铱莹,彭其渊.客运专线对运输通道分担率的影响[J].铁道运输经济研究,2006,28(12)：16-19
[21]何宇强,毛保华,陈团生,杨静.高速客运专线客流分担率模型及其应用研究[J].铁道学报,2006,28(3)
[22]曾曦,王慈光.Logit模型的改进及其在客流分担率预测中的应用——成渝城际铁路为例[J].长沙交通学院学报,2007,23(4)：50-53
[23]张戎,吴晓磊,张天然.基于RP/SP融合数据的沪杭客运通道公铁客流分担率研究[J].铁道学报,2008,30(3)：7-13
[24]张戎,吴晓磊.长三角快速客运网城际公共交通分担率模型[J].同济大学学报(自然科学版),2008,36(9)：1211-1215
[25]吴才锐,过秀成,何明,姜晓红,陈永茂.基于广义费用修正模型的客运方式选择研究[J].交通与计算机,2008,26(4)：132-134
[26]王爽,赵鹏.基于Logit模型的客运专线旅客选择行为分析[J].铁道学报,2009,31(3)：6-10
[27]李滨汀,陈子君,王英.京津城际动车组对旅客出行方式影响分析[J].交通科技与经济,2010,0 2
[28]张力.基于广义费用的客运专线客运需求预测方法研究[D].北京：北京交通大学,2010
[29]张涛.区域通道内旅客出行方式选择研究[D].成都：西南交通大学,2004
[30]关宏志.非集计模型-交通行为分析的工具[M].北京：人民交通出版社,2004：13-26
[31]刘灿齐.现代交通规划学[M].北京：人民交通出版社,2001：140
[32][美]迈克·D·迈耶,埃里克·J·米勒.城市交通规划[M].第二版.杨孝宽.北京：中国建筑工业出版社,2008：198-207
[33]Wikipedia. Discrete Choice [EB/OL], http://en.wikipedia.org/wiki/Discrete_choice, 2011-04-05
[34]宋素娟.燃油价格对小汽车出行需求的影响研究[D].北京：北京工业大学,2010
[35]Simon P.Washington. Matthew G.Karlaftis. Fred L. Mannering. Statistical and econometric methods for transportation data analysis[M].Cahpman & Hall/CRC,2003
[36]Juan de Dios Ortuzar, Luis B. Willumsen. Modeling Transport [M]. third edition. JOHN WILLY & SONS, LTD.
[37]姜学方.基于行为的交通出行预测[D].西安：西安建筑科技大学,2006
[38]邵昀泓,王炜,程琳.出行方式决策的随机效用模型研究[J].公路交通科技,2006,23(8)：110-115
[39]王静龙,梁小筠.定性数据分析[M].上海：华东师范大学出版社,2005：141-154
[40]米红,张文琦.实用现代统计分析方法与SPSS应用[M].北京：当代中国出版社,2004
[41]金宝辉.交通出行行为分析[D].成都：西南交通大学,2004
[42]火车票车次、里程[EB/OL]..http://huoche.kuxun.cn/shike-wenzhounan.html
[43]汽车票价及里程查询[EB/OL].http://www.piaojia.cn/changtuzhan/z362/
[44]张曾种.道路运输通道交通特性研究[D].西安：长安大学,2002
[45]白惠涛.基于多种交通方式的城市快捷客运系统出行路径选择模型及优化方法研究[D].北京：北京交通大学,2007
[46]戴玲玲.高速铁路与其他运输方式竞争关系研究——以京沪线为例[D].北京：北京交通大学,2009
[47]郑璐.公路客运与城际铁路协同发展研究[D].长春：吉林大学,2008
[48]李群仁,曾会欣,张力.京沪走廊各种运输方式的社会成本[J].中国铁路,1999,6：6-9
[49]高铁开通对浙江省公路运输的影响分析[R].杭州：浙江公路水运工程咨询公司、长安大学,2010
[50]富振英.Logistic回归模型数据的预处理(二)——离散变量数量化方法[J].中国卫生统计,1992,9(2)：43-45
[2]高速时代风起云涌公路客运何去何从.交通世界,2009年第12期.
[3]中长期铁路网规划(2008年调整)。
[4]Benedikt Mandel, Marc Gaudry, Werner Rothengatter. (1994) linear or nonlinear utility functions in logit models? The impact on German high-speed rail demand forecasts [J].Transpn.Res.28B (2),91-101
[5]PETER NIJKAMP, AURA REGGIANI, TOMMASO TRITAPEPE:MODELLING INTER-URBAN TRANSPORT FLOWS IN ITALY:A COMPARISON BETWEEN NEURAL NETWORK ANALYSIS AND LOGIT ANALYSIS [J]. Trans Res.-C, Vol.4, No.6, pp.323-338,1996
[6]Agostino Nuzzolo, Umberto Crisalli, Francesca Gangemi. A behavioural choice model for the evaluation of railway supply and pricing policies[J]. Transportation Research Part A,2000, V(34):395-404
[7]Moshe Ben-Akiva, Takayuki Morikawa. Comparing ridership attraction of rail and bus[J]. Transport policy,2002, V9:107-116
[8]Enjian Yao T. Morikawa, S. Kurauchi, T. Tokida. A study on nested logit mode choice model for intercity high-speed rail system with combined RP/SP data [J]. Proceedings of the Conference on Traffic and Transportation Studies, ICTTS, p 612-619,2002
[9]Gonzalez-Savignat, Mar (2004)'Will the high-speed train compete against the private vehicle?',Transport Reviews,24:3,293—316
[10]Chang, Iljoon and Chang, Gang-len(2004)'A network-based model for estimating the market share of a new high-speed rail system', Transportation Planning and Technology, 27:2,67—90
[11]Yonghwa Park, Hun-Koo Ha. Analysis of the impact of high-speed railroad service on air transport demand[J].Transportation Research Part E 42 (2006) 95-104
[12]石定宇.运输方式选择行为模型的参数估计方法[J].西南交通大学学报,1991,4：1-7
[13]石定宇,高旭敏.运输通道中公路与铁路旅客选择行为模型研究[J].铁道学报,1992,14(2)：58-63
[14]杨兆升,黎春兰.客运方式选择方法研究[J].中国公路学报,1997,10(2)
[15]黄承锋,杨秀苔,李毅.运输通道市场结构变动的静态博弈分析——以成渝通道客运为例[J].重庆交通学院学报,2000,19(4)：105-109
[16]胡文友,李旭宏.公路运输通道内的车辆出行路径选择模型及应用研究[J].公路交通科技,2003,20(1)：125-128
[17]马波涛,张于心,赵翠霞.运用Logit模型对高速客流分担率的估计[J].北方交通大学学报,2003,27(2)：66-69
[18]孙海林,胡兆杰.秦沈客运专线对京沈客流分担率的影响分析[J].铁道运输与经济研究与建议,2004,26(3)：69-71
[19]谢如鹤,邱祝强,李庆云,王荣华.Logit模型在广深铁路客流分担率估算中的应用[J].中国铁道科学,2006,27(3)：111-114
[20]张铱莹,彭其渊.客运专线对运输通道分担率的影响[J].铁道运输经济研究,2006,28(12)：16-19
[21]何宇强,毛保华,陈团生,杨静.高速客运专线客流分担率模型及其应用研究[J].铁道学报,2006,28(3)
[22]曾曦,王慈光.Logit模型的改进及其在客流分担率预测中的应用——成渝城际铁路为例[J].长沙交通学院学报,2007,23(4)：50-53
[23]张戎,吴晓磊,张天然.基于RP/SP融合数据的沪杭客运通道公铁客流分担率研究[J].铁道学报,2008,30(3)：7-13
[24]张戎,吴晓磊.长三角快速客运网城际公共交通分担率模型[J].同济大学学报(自然科学版),2008,36(9)：1211-1215
[25]吴才锐,过秀成,何明,姜晓红,陈永茂.基于广义费用修正模型的客运方式选择研究[J].交通与计算机,2008,26(4)：132-134
[26]王爽,赵鹏.基于Logit模型的客运专线旅客选择行为分析[J].铁道学报,2009,31(3)：6-10
[27]李滨汀,陈子君,王英.京津城际动车组对旅客出行方式影响分析[J].交通科技与经济,2010,0 2
[28]张力.基于广义费用的客运专线客运需求预测方法研究[D].北京：北京交通大学,2010
[29]张涛.区域通道内旅客出行方式选择研究[D].成都：西南交通大学,2004
[30]关宏志.非集计模型-交通行为分析的工具[M].北京：人民交通出版社,2004：13-26
[31]刘灿齐.现代交通规划学[M].北京：人民交通出版社,2001：140
[32][美]迈克·D·迈耶,埃里克·J·米勒.城市交通规划[M].第二版.杨孝宽.北京：中国建筑工业出版社,2008：198-207
[33]Wikipedia. Discrete Choice [EB/OL], http://en.wikipedia.org/wiki/Discrete_choice, 2011-04-05
[34]宋素娟.燃油价格对小汽车出行需求的影响研究[D].北京：北京工业大学,2010
[35]Simon P.Washington. Matthew G.Karlaftis. Fred L. Mannering. Statistical and econometric methods for transportation data analysis[M].Cahpman & Hall/CRC,2003
[36]Juan de Dios Ortuzar, Luis B. Willumsen. Modeling Transport [M]. third edition. JOHN WILLY & SONS, LTD.
[37]姜学方.基于行为的交通出行预测[D].西安：西安建筑科技大学,2006
[38]邵昀泓,王炜,程琳.出行方式决策的随机效用模型研究[J].公路交通科技,2006,23(8)：110-115
[39]王静龙,梁小筠.定性数据分析[M].上海：华东师范大学出版社,2005：141-154
[40]米红,张文琦.实用现代统计分析方法与SPSS应用[M].北京：当代中国出版社,2004
[41]金宝辉.交通出行行为分析[D].成都：西南交通大学,2004
[42]火车票车次、里程[EB/OL]..http://huoche.kuxun.cn/shike-wenzhounan.html
[43]汽车票价及里程查询[EB/OL].http://www.piaojia.cn/changtuzhan/z362/
[44]张曾种.道路运输通道交通特性研究[D].西安：长安大学,2002
[45]白惠涛.基于多种交通方式的城市快捷客运系统出行路径选择模型及优化方法研究[D].北京：北京交通大学,2007
[46]戴玲玲.高速铁路与其他运输方式竞争关系研究——以京沪线为例[D].北京：北京交通大学,2009
[47]郑璐.公路客运与城际铁路协同发展研究[D].长春：吉林大学,2008
[48]李群仁,曾会欣,张力.京沪走廊各种运输方式的社会成本[J].中国铁路,1999,6：6-9
[49]高铁开通对浙江省公路运输的影响分析[R].杭州：浙江公路水运工程咨询公司、长安大学,2010
[50]富振英.Logistic回归模型数据的预处理(二)——离散变量数量化方法[J].中国卫生统计,1992,9(2)：43-45