Methodology review

There is a significant lack of studies in the field of route planning for railroads. Usually scientists evaluate given alternatives or make some specific studies about the project that already exists, but there are no commonly used methods, for the case, when the routes should be generated first. Nevertheless, some common methods of transportation will be used.

Regarding transportation routes evaluation, the most widespread are the studies devoted to demand estimation, assessment of external effects, such as air pollution, noise and vibration pollution, and influence on living areas.

The most generally-used model for demand estimation is transportation demand analysis (TDA). It consists of 4 stages: trip generation, trip distribution, modal shift, and traffic assignment.

According to Caulfield (2011), trip generation refers to the amount of trips generated by each origin and destination that are influenced by several socio-economic feature of the zone. It includes two components:

 trip production (by the origin points), can be influenced by population, wages on the macro-level and by family size, social status, availability of car on the household level;

 trip attraction (by destinations) is usually assumed to be influenced by the type of land use, economic activity, employment.

Trip distribution, as a second step of TDA, may be formed in different ways, but the most

widely used is gravity approach to the distribution of trips. It assumes that trips from each origin towards each destination are distributed according to Newton’s gravity law as follows:

)

f(c

_ij

): generalized function of travel cost between i and j.

Generally speaking f(c_ij

) here is a resistance factor, so not only the cost can be used at this

place, but also travel time or distance.

Mode split is usually based on the utility of each alternative mode that consists of predicable

value V and random value ε as follows:

m

13

v

_m: predicable utility component of mode m;

ε

m: random utility component of mode m.

Predictable component is a function of the transportation mode characteristics, such as travel cost, travel time, additional costs, etc. Finally, using the utilities calculated, mode choice model is built as a multinomial logit model as follows:

 

v

m: predicable utility component of mode m.

The last step of the conventional model – traffic assignment – concerns link availability as a supply and number of O-D pairs and transportation modes as a demand. This step operates with a term of Level of Service (LOS) for each link of the network: the higher is the demand on each link, the worse are the traffic conditions and the larger is travel time. Mathematically:

QS

e

V

t t 

₀ ^/

where

t: travel time at the link;

t

0: travel time under free flow conditions;

V: flow;

Q

s: link capacity.

It should be taken into account that transferring an O-D pair from one link to another will simultaneously cause improvement of LOS on the former, and decline of LOS on the latter.

According to the purposes of analysis, this method can be used partially, depending on the detailing of analysis needed. Also, TDA can be used in several variations, depending on the data available. For example, Ehrenbreger et al. (2010) uses a TDA that joins stages 1 and 2 of the analysis for European HSR lines. This integrated model utilizes socio-economic data of two cities and passenger traffic on the link between them. The study develops a gravity model for European cities, where transportation demand can be estimated via GDP, tourism intensity, population size, and distance. In the final model tourism data is omitted, because it does not influence result.

where F_ij: travel amount between cities i and j;

(3)

(4)

(5)

14 β0… β3: coefficients;

P_i, P_j: population of cities i and j;

Wi, Wj: GDP of cities i and j;

dij: distance between cities i and j;

vij: average speed on the link between i and j.

We should note that although the above study is devoted to the development of second-generation HSR, it offers methods, useful for this research.

Evaluation of HSR route requires not only travel demand data, but also information about engineering and environmental feasibility of the project. Depending on the accuracy level needed and available data, methods can vary from rather precise monetary evaluation to rough assessment, but generally they usually use Geographic Information Systems (GIS).

Both kinds of feasibility refer to the construction complexity, and, consequently, to construction cost. According to Uršej and Kontić (2007), it is influenced by such factors:

 cost of special constructions, such as embankments, cuttings, bridges and tunnels;

 cost of living-areas protection, such as noise and visual barriers;

 cost of natural and cultural heritage protection: remediation of environment, creating crossings for animals;

 additional cost, such as related to land purchase.

The key idea of route assessment in the study of Uršej and Kontić (2007) is thesis, that HSR should go underground, only in the cases, when surface solution is unavailable due to surface space, causes great negative environmental impacts, or is more expensive than subsurface alternatives. That’s why route planning starts form the suitability analysis of the surface where the alternative route should lie. Finally, alternative routes are ranked with respect to length of the route, number of tunnels and subsurface sections length.

GIS is also used in the study by Ehrenbreger et al. (2010) to evaluate technical complexity of the future line. This estimation is based on two ideas: there is a basic construction cost per kilometer and resistance multipliers that are calculated from resistance maps. At the resistance map, each raster pixel is assigned a resistance value depending on geographic conditions.

Parameters that significantly influence construction cost are:

 Slope of the terrain.

As railway lines are restricted to the maximum gradient (for HSR maximum is usually 35-40‰ on the exclusive tracks, (UIC, 2010)), the cases exceeding this limit have to be handled by building of embankments or tunnels, bridges. Bridges and tunnels are assumed to be 7 times more expensive than the basic cost value.

 Areas with high population density.

15

A higher population density leads to the higher construction costs, because such areas have a few free areas for HSR line. So, the population density over 100 persons per pixel causes linear growth of the resistance multiplier. If the population is less than 100 persons per pixel, it does not affect construction cost.

 Water bodies.

Crossing of rivers is assign to have a resistance factor of 5.4, other water bodies cannot be crossed except the British channel. As it already has a tunnel, it has a small resistance factor of 1.5 (larger than 1 because of high fares).

More criteria for route evaluation can be found in California High Speed Rail Authority (2012) that develops HSR route in California, USA. One of the issues studied is a section of HSR between two cities and environmental impact of alternative routes. To evaluate the routes, two groups of criteria are introduced: physical and operational characteristics (such as travel time (minimization), intermodal connections (maximize), route length (minimize) etc.) and environmental impacts (air pollution, noise, vibrations, cultural and human hazards). Most of these criteria are already discussed, but there is one extra – intermodal connections, that helps to evaluate HSR line not only as a single transportation unit, but a part of transportation infrastructure of the region.

16

在文檔中 新興國家高速鐵路路線：以烏克蘭為例 (頁 22-26)