? is given control in modifying or changing the

? Provides a cost effective way
of testing and evaluating different scenarios ? Allows the user to test
scenarios faster than in real life ? Offers an insight into the
characteristics of traffic system operations that are important, allowing the
user to make a more informed decision ? Provides outputs/animation
that the public can understand At the same time simulation has some
disadvantages: ? Requires collection of
detailed data on field conditions ? Needs calibration and
validation of the model prior to testing scenarios ?
Requires an understanding of how the model works before assessing the outputs
The processes of conducting a simulation study are described in the following
section. : The movement of vehicles in a microscopic simulation program is
controlled by different logic. There might be logics for car-following, lane
changing, merging, behavior on freeways and urban routes. It is necessary to
have a basic understanding of the logic being applied in the simulation model
before selecting it. It is also helpful to know if the user is given control in
modifying or changing the logic in the model. Ease of use: Each simulation
model is different in terms of its ease of use. Some simulation models provide
an easy to use graphic user interface (GUI) and have a simple process for data
input. Some models allow the user to use an aerial photograph or map as a
background. while coding the network, which makes geometry coding much easier.
And some models are easy to calibrate/validate by providing different measures
of effectiveness in a readable format. Simulation models provide users with
animation of the network. Some models have the capability to provide 3-
dimensional animations with other details (like buildings, traffic signals,
etc.) and also to allow the user to zoom or pan across networks. Most of these
animations can be made into movie clips and are great tools for presentations,
providing a vivid presentation of the benefits of TSP to policy boards and
public. Most simulation models have a provision that allows for customizing
output format. It helps the user to have the results in a compiled form rather
than a vehicle-by-vehicle or a link-by-link basis. Sometimes the models also
support tracking individual vehicles and this might be useful in the case of
transit signal priority. Travel time, delay, queue length, number of stops,
fuel consumptions, emissions, bus/tram wait time are some of the most used
outputs. Field data is required to conduct a simulation study. The data usually
required to conduct a simulation is as follows: ? Geometric — length of lanes,
lane widths, number of lanes, lane assignment, length and location of
crosswalks ? Signal timing plans – cycle lengths, number of
phases, duration of phases, etc. for different times of the day ?
Speed – speed limits and average vehicular speed ? Transit vehicle – travel
times, dwell times and distribution of dwell times The actual amount of data
collected depends on available budget, time and project requirements. At a
minimum a.m. and p.m. peak data should be collected. If there is a significant
variation in traffic volumes, additional data may need to be collected such as
midday, Saturday and Sunday. Attention should be given to the time period of
data collection. For example, data collection of regular weekday traffic should
be conducted between Tuesday and Thursday to avoid traffic fluctuations due to
weekend trips and preferably, should not be during a holiday season. Weekend
data should be collected during times that are most relevant to meet project
objectives, such as weekend traffic peak period and/or weekend peak transit
operational headways



Developing the network for simulation involves
the processes of coding, calibration and validation. Calibration and validation
are extremely important, but often do not get the attention they deserve in the
simulation process. Coding: Coding is the process of building the network in
the simulation model. It involves input of the study area geometry in the
simulation model with other information like the timing plans and volumes.
Given below are the different ways to code a network in a simulation model: ?
The network can be coded in the simulation model itself. ?
The network can be coded in a Network Editor that is bundled with the
simulation model and can then be transferred to the simulation model. The
Network Editor is designed to simplify network coding. ? In
some cases the network can be coded in simulation or optimization models and
can be transferred/exported to other simulation models. This procedure is
mainly used when multiple softwares are required for the analysis and
optimization of the network. Sometimes, due to the simplicity of the Network
Editor, the study network is coded in one model and then transferred to another
model. If this method of coding the network is used, the user should be aware
of the features/elements that may be altered or missed during the transfer
process. Calibration: Simulation models need to be calibrated before they are
used for testing future scenarios since an un-calibrated model might lead to
skewed results. Calibration involves changing some of the default parameters
used in the simulation model so that it reflects the conditions observed in the
study area. As previously mentioned, the movement of traffic in a simulation
model is governed by different logics. Simulation models provide the user with
different parameters to fine tune the logic and other aspects of the model.
Examples of these parameters include minimum headway, minimum acceleration
rate, minimum deceleration rate, maximum acceleration rate, etc. These parameters
are give

We Will Write a Custom Essay Specifically
For You For Only $13.90/page!

order now