|
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
 |
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
BUDAPEST METRO LINE 4 FEASIBILITY STUDY Oktober 1996 |
 |
|
Transport model |
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Introduction
The basis of the demand forecasts for the study is
a public transport model of Budapest which has been constructed
and tailored to the needs of the study. The model contains the
whole of Budapest with particular emphasis on the south Buda corridor
and the trips of interest, namely those between South Buda and
the Pest CBD. In developing the base year model, the study team
concentrated on reproducing the current service levels and the
current passenger demand on the network, reflecting sub-modal
utilities and the split of demand between bus and tram services
on an assignment basis. A reasonably accurate reproduction of
current demand levels and their sub-modal split will ensure a
sound basis for the forecasting work. The model was constructed
on an all day basis reflecting the available demand data used
as the basis for developing the demand matrix for the study.
As peak demand is required for system design purposes, factors
were developed to convert all day demand to peak demand at a link
level for relevant sections of the network. This chapter reports on the work undertaken to construct and calibrate the base public transport model which will be used for the demand forecasting. Sources of dataData sources used for the construction and calibration of the model were listed in the above paragraph 2.5.1. Network and details of servicesPhysical network
Details of the physical network which describe Public
Transport services are coded in a network data file. This file
contains the following physical details used by the model:
Figure 3.1 shows the network within the study area. Lines/Service descriptionsFollowing the definition of the physical network, public transport services which run on the network are defined in detail in a data file referred to as the lines file. There are some 350 route descriptions included in the lines file. This file contains the following data for each line:
The network and lines files work together to define the full characteristics of each service. All the above data were checked through sampling on the network and reference to independent sources. Fare modelA fare model is included within the overall Public Transport model developed for the study. The effect of the fare model in the base year is rather limited in that it does not play a role in the routeing of passengers. This is because the fare regime in Budapest is based on a flat fare system and almost all regular users of the system hold a Public Transport pass. This means that the decision on the route to choose for a journey is not dependant on the fare paid as cost of all journeys for pass holders is identical. It is therefore not foreseen that the fare model will play a significant role in the utilisation of services and route choice. However, incorporation of the fare model will provide an efficient mechanism for checking the calculation of total revenues. The fare model incorporates a zone system with two zones. One zone for Budapest (within the city boundary) and the other for the agglomerations. The fare levels were calculated as an average based on total fare revenues by ticket type and total passengers using each ticket type in 1995. The analysis for the agglomeration only included those in the south west of the city as only these are of interest to the study. Demand matrix
The demand matrix was constructed and calibrated
in two stages as follows:
The demand data available for the construction of
the matrix was as follows:
In the first instance, based on the above data sources
an overall matrix was constructed. This matrix was then put through
a matrix estimation process to reproduce total, all modes, passenger
counts across screenlines and on individual links throughout Budapest
with specific reference to the south west corridor. Over 90 passenger
loading count sites were input to the process. Following a number
of logic checks and estimation runs, the matrix estimation process
resulted in a matrix which was considered suitable for use with
the study. This matrix was constructed based on the 164 zones
of the Budapest General Transport Surveys. Zones outside of the
corridor of interest were aggregated and zones within the area
of interest were kept as the 164 zoning system. After aggregation
the matrix contained 87 zones. Following the construction of the above matrix, zones in the corridor of interest were disaggregated to allow modelling of passengers boarding or alighting at stops in more detail. Disaggregation of zones were based on population and employment data on a much finer Urban Planning Zones system and below this level, based on site visits and local knowledge of the area. The resulting disaggregated matrix contains 298 zones which was the principal input to the assignment model and was used in the validation process. Model CalibrationAssignment Specification
An assignment model was tested which contained the
298 zone matrix and lines/services and networks files as described
above. The assignment was based on a multi-routeing algorithm.
The robustness of the assignment method was tested through a
number of sensitivity tests which tested the sub-mode split model
for boarding at origin, interchange and spread of the routes for
inclusion within the range of routes which are considered for
assigning trips for each Origin-Destination pair. The model was
generally found to be very stable with resulting changes corresponding
to changes in the input parameters and proportionate to the size
of change in these parameter, within an expected range.
Each path contains the following components:
There is a factor attached to each of the above components
within the assignment model. These factors are combined within
a generalised cost formula of the form below: Generalised Cost = (Access time *
Wa + Wait time * Ww + Transfer time * Wt + Transfer penalty * Nt+ Boarding penalty*Nb)
+ (In-vehicle time * Wmi) + F where: Wa = weighting on access/egress time values.
The assumptions adopted for each component of the generalised cost function are described in detail in Stage 1 Report.. Checks between zone paths
Checks were undertaken on whole paths, between origins
and destinations, to ensure that the overall path building process
works correctly and all components are incorporated as excepted.
These checks were based on a manual calculation of generalised
time costs for a sample of paths built by the model and confirmed
the model's internal calculations to be working correctly. The
majority of the sample paths which were checked were within the
study corridor, namely for trips between District XI and Districts
V, VI, VII, VIII and IX. Other paths between south west agglomerations
and CBD Districts as well as paths from District XI and XXII and
District I were also checked to ensure reasonable paths are produced
by the model. The area of interest contains over 100 model zones which results in over 10,000 main (or primary) transport paths and up to some 20,000 to 30,000 secondary paths. From these some 100 main paths were examined all of which displayed reasonable routeings. Furthermore, sensitivity tests and consistency in model results in the comparison of observed against modelled data indicated reasonable and stable routeings. Validation - Comparison of observed and modelled passenger counts
Passenger flows produced by the model were compared
at different levels against the observed data. The most widely
reported comparisons are generally as follows:
The most detailed level of data available for this study was
the number of passengers by mode for each link on the network. Tables 3.2 and
3.3 contain the results of the comparisons across screenlines and at scattered
individual sites, respectively. Table 3.1 shows that in overall terms, on a
screen line across the Danube, the modelled flows exceed the observed flows
by some 2% with the bus mode over-predicting by 5% and the tram mode under-predicting
by 3%. Given the variability in the observed data and the sampling techniques
used in deriving the modelled data we consider that the differences are insignificant.
Table 3.2 shows that for individual sites modelled and observed flows are within
10% except at one site. For individual sites, this is considered acceptable. From our analysis we conclude that the demand model validates well and is suitable to be taken forward to the forecasting stage. Highway model
A capacity restrained urban highway model, using
SATURN software, which covers the whole of Budapest has been
made available to the study team for undertaking comparative analysis
of the effects of the public transport options on the highway
network. The model includes all junctions within the central
area of Budapest and contains speed flow curve data for links
outside of the downtown area. The highway traffic demand data
used with this model is based on the data from the Budapest General
Transport Surveys (1992-94). This model is currently being used
for a number of studies within Budapest. Our examination of this
model and the results produced from it both on the base year network
and various test networks showed that the model was consistent
and stable and it is therefore considered suitable for use in
this study for comparative assessments.
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||