VISSIM 4.1 is now being released in April 2005 and its main new features like additional COM-functions, a parking model, a new interface for fixed time signal control and the snapshot feature to start from pre-simulated scenarios will be presented. Currently the strategic demand modeling software VISUM 9.2 is being distributed with more flexibility to plot networks, improved GIS-like network editing and extended junction modeling. In July 2005 VISUM 9.3 will be launched; some of the new ideas will be shared at the presentation.

Return on top

Dynamic Automation in Disaggregate Modelling using VISUM and VBA for PRISM West Midlands

By Rajat Bose (Mott MacDonald) and Sonal Ahuja (Mott MacDonald)

Email: rajat.bose?mottmac.com, sonal.ahuja?mottmac.com (please substitute ‘?’ with ‘@’ – spam preventing measure)

Abstract:

PRISM (Policy Responsive Integrated Strategy Model) is a travel demand modelling system developed for the West Midlands. The aim of the PRISM model is to forecast all travel responses (Highway and Public Transport) to a variety of policies and proposals for the region, and demonstrate their impact at a regional level.

With the PRISM model, it is required to test the impacts of these policies and proposals, by conducting a number of iterations of the model between the network component developed in VISUM, and the demand component, developed using ALOGIT. For each of these, the population forecasts, assignments, skim generation, and travel demand estimation processes are time consuming; the network has a total of 898 zones and 25,000 links, plus 6,000 explicitly coded junctions. Doing this process manually would require a great input of time and effort, whilst this would also preclude efficient convergence monitoring. Therefore this required an automated procedure to carry out this sequential process. The advantage we had here was that both the VISUM and the DOS controlled ALOGIT files can be controlled using VBA. Therefore this platform was used to create a shell for managing the iterative process.

This paper presents the process and structure of developing the automated interface for PRISM West Midlands. This includes incorporating the sequential structure for running the demand model, where outputs from one stage work as the inputs for the other.

A fully automated interface has been developed using VBA, and the COM interface of VISUM, which links VISUM (assignment model) to ALOGIT (Demand Model), MUULI (the matrix manipulation model) and MS Office (model output analysis tool). When developing the interface, it was required to identify how the users would wish to operate the model, understand what outputs are required and how different parts of the model relate to each other. This included developing a viable file structure, which would be easily interpreted by the user; also ensure the system would allow the user to operate the processes from various stages of the model. This is because it is not always required to run the entire process to test the alternatives, policies and programmes. This meant that the shell had to be dynamic enough to be able to operate from different starting points, where the new runs would require only partial input data, while the rest would use results already generated from various earlier stages.

As the PRISM model is a disaggregate model, the model is generating travel demand for 10 purposes. Based on these, peak period and peak hour matrices are generated for the AM, PM, inter peak and off-peak periods for 4 highway purposes and the 3 PT modes. The PRISM integration model has been developed to operate on two levels. The first level is where population forecast data is generated from the planning data. This is followed by loading the reference year and scenario network with the base year matrix and running through the model to generate the first set of synthetic matrices. The next level is to load these synthetic matrices on the network and repeat the process for ‘n’ number of iterations. Currently 3 iterations are done to get a stable set of demand matrices. However, there is a possibility of introducing more explicit convergence criteria, which may be added to the process.

Following the model run, it is possible to analysis the results in terms of the modes used, purposes for which the trips are made, cross mode-purpose analysis, etc. Increased automation is proposed to be achieved by using COM interface to produce a standardised output and post processing of data for analysis model results.

Return on top

Accessing VISUM objects through the COM interface

By Charles Lindveld (Imperial College London), Steffen Weckek (PTV)

Email: k.lindveld?imperial.ac.uk (please substitute ‘?’ with ‘@’ – spam preventing measure)

Abstract

A perhaps under-appreciated aspect of the VISUM package is that it allows external software full access to all of its capabilities via the COM interface. What is even less well-known is that in Microsoft Excel almost all Windows users have very powerful tools to use this COM interface. This is because Excel includes Visual Basic for Applications (VBA), a simple but powerful programming language.

In other words: all data stored in VISUM files (networks, O-D matrices, Skim matrices, modal split etc., and all VISUM procedures are available from VBA scripts that can be created and run from Microsoft Excel.

We have used this mechanism to automatically extract network link-path incidence matrices with links from VISUM. Link-path matrices list all paths through the network that are being used by traffic, plus which links are on each path.

Being a self-contained environment for network manipulation, involving routes and paths, and being conceived for practical application rather than research, VISUM does not contain primitives to dump link-path incidence matrices. However, direct access to the objects in which VISUM stores its data permits one to use the functionality already available from within VISUM to access and output such matrices using a VBA script.

In the presentation the working of the VBA script will be shown interactively on a simple network.

Return on top

An application of VAP: modelling the interaction between vehicles travelling on opposing links and simulating reaction to Variable Message Signs

By Giacomo Tuffanelli (JMP)

Email: Giacomo.Tuffanelli?jmp.co.uk (please substitute ‘?’ with ‘@’ – spam preventing measure)

Abstract:

JMP Consulting were commissioned by the London Borough of Richmond upon Thames to investigate the feasibility of a Variable Message Signing scheme to control movements of large vehicles on a section of Petersham Road, characterised by tight corners and narrow carriageway widths.

In this particular case-study the main cause for delays to general traffic is the difficulty encountered by large vehicles travelling in opposite directions through the narrow sections of the road. This often results in opposing vehicles coming to a virtual standstill with frequent cases of HGVs mounting the kerbs presenting an added safety hazard to pedestrians. The modelling of driver behaviour when faced with these situations has been achieved through the use of VISSIM’s inbuilt VAP module.

VAP recognizes and controls key elements of the network relating to vehicle behaviour. In this case, a series of detectors and desired speed decision sections were used to achieve the modelling of the base scenario. A virtual signal controller was set up to link the various elements to the VAP file. The detectors were used to monitor the presence of vehicles on opposing links, while the desired speed decision sections were used to adjust the speed of traffic depending on the presence of oncoming vehicles . For the proposed model, a similar technique has been employed, however this time the detectors enable vehicles to recognise the indication given by the VMS sign and yield to oncoming traffic.

The model outputs were used to estimate improvements to general traffic and bus operations deriving from the implementation of the VMS scheme. The results show that, if the VMS sign is adequately observed, then there should be a journey time benefit along Petersham Road. LB Richmond upon Thames are considering implementing the scheme on an experimental basis and monitoring the results. This will also add a valuable insight into driver reactions to VMS signs, a topic that has not been widely researched.

Return on top

Evaluation of the Impact of Highway Schemes on Pedestrians using VISSIM

By Paul Clifford (Jacobs Babtie)

Email: paul.clifford?jacobs.com (please substitute ‘?’ with ‘@’ – spam preventing measure)

Abstract:

Jacobs Babtie has been undertaking research into the application of micro-simulation models for all road users, including pedestrians, to assess the ability of such models to evaluate options for all users on an equitable basis. The study, commissioned by Transport for London, was undertaken jointly with Imperial College to develop new procedures. The deliverables have included the development of a multi-criteria evaluation framework and option ranking system to enable transport planners to maximize the output of micro-simulation models when comparing potential options at junctions/interchanges. The techniques drawn together and the processes developed provide the potential for transport access and equity issues to be considered in a logical manner and at an early stage in option development.

Return on top

Modeling Pedestrian Flow in VISSIM: Possibilities and Limitations

By Muhammad Moazzam Ishaque (Imperial College London)

Email: m.ishaque?imperial.ac.uk (please substitute ‘?’ with ‘@’ – spam preventing measure)

Abstract

VISSIM offers two different ways to model pedestrian flow. The first of these is to specify no-interaction between pedestrians. Under such a constraint all pedestrian movements are independent of the presence of other pedestrians in the vicinity. Any speed flow relationship is essentially linear with speed remaining equal to the desired pedestrian speed irrespective of the pedestrian flow and density. The second possible way to model pedestrian flow is to use Wiedemann’s Vehicle Following Model. In such a situation pedestrians react to the presence of other pedestrians, although under the rules developed for vehicles rather than pedestrians. In this modeling environment there is a possibility to achieve a speed flow and speed density relationship for which calibration can be attempted. This paper reports on some simple experiments performed in VISSIM to model pedestrians under different flow and density conditions and different simulation parameters. The aim is to show the best possible modeling options given the limitations of the software.

Return on top

Application of Micro-Simulation to Strategic Motorway Planning - M1/M62 Interchange: A Combined VISSIM-TRANSYT Assessment

By David Keenan (Faber Maunsell)

Email: david.keenan?fabermaunsell.com (please substitute ‘?’ with ‘@’ – spam preventing measure)

Abstract:

In 2004, FaberMaunsell, under the newly established “Network Analyst” role in the Highways Agency’s North of England region, carried out a VISSIM micro-simulation study of perceived “un-necessary” network congestion in the vicinity of the M62 – M1 Lofthouse Interchange near Leeds, West Yorkshire in response to a complaint received from the Freight Transport Association by the Network Strategy and Traffic Operations Directorates of the Highways Agency. The aim of the project was to identify low-cost "early-win" schemes which could be implemented quickly and would compliment the £1.2bn motorway interventions due to be built in 2006 - 12 under the South and West Yorkshire Motorway Best Use Study (SWYMBUS). Innovative designs, intended to maximise capacity of the existing available road space were developed and shown using VISSIM micro-simulation models to have significant benefits for motorway users in terms of eliminating existing congestion problems by achieving better segregation of the motorway traffic streams.

To limit scheme costs, intervention proposals considered under the project were restricted to revisions to white-lining and gantry signing to achieve more efficient motorway lane-usage and signal timing improvements at the interchange roundabout to optimise junction throughputs, which necessitated inputs to the VISSIM traffic signal simulations from a linked TRANSYT model of the associated grade-separated signalised roundabout. In addition to the more usual numerical analysis, 3-D visualisation techniques were utilised to “fly-through” the interchange and illustrate the potential improvements to traffic congestion possible in advance of scheme implementation and construction. This paper details the methodology adopted in the study and summarises the results of the intervention-analysis performed.

Return on top

Airside model of the New Bangkok Airport using VISSIM

By Philippe Perret (Scott Wilson) and Timmy Chan (Scott Wilson)

Email: philippe.perret?scottwilson.com, timmy.chan?scottwilson.com (please substitute ‘?’ with ‘@’ – spam preventing measure)

Abstract:

This new airport will take over the functionality of the old airport in September 2005 and is aspiring to be one of the major hubs in South-East Asia. It is predicted that this new airport will cater for 54 MPPA. The client wanted to test the efficiency of the airside network for future demand, and identify pinch points in the network, so that any changes to the design can be accommodated. The main focus of the investigation centered on the area connecting the main airport site with a new satellite stand. These areas would be connected via an underground Tunnel. There was real concern as to whether one tunnel would be sufficient, whether the priority junctions would operate satisfactorily, and whether more points of access were required.

This modelling involved the representation of almost 150 aircraft stands including provisions for the new A380, together with the landside and taxiway networks. Numerous vehicle types were also considered for the model to simulate movements of baggage, maintenance, fuel, and passenger buses. Information from SIMMOD, a package used to simulate flight scheduling, was fed into VISUM and VISSIM to generate a network covering 70km. This project represents an expansion of the modelling capabilities of VISSIM.

Return on top

Tips from the Hotline - the most common modeling errors in PTV VISION

By Dr. Martin Fellendorf (PTV), Karsten McFarland (PTV)

Email: martin.fellendorf?ptv.de , karsten.mcfarland?ptv.de (please substitute ‘?’ with ‘@’ – spam preventing measure)

Abstract:

The PTV hotline is getting all kinds of questions from beginners and experienced users as well. Some of these questions are repititive and indicate a misunderstanding of the modeling concepts with VISUM and VISSIM. This presentation will highlight some of these most common misunderstandings. For example in VISSIM it is essential to use single connectors even for links with multiple lanes. Furthermore the usage of link types is essential in both packages. Tips like these will be presented using bad and good examples.

Return on top

Efficacy of Dynamic Route Guidance Systems – A micro-simulative approach

By Thogulava Hariharan (Atkins) and Sergio Grosso (TORG, PTV)

Email: hariharan.thogulava?atkinsglobal.com, sergio.grosso?ptv-newcastle.co.uk (please substitute ‘?’ with ‘@’ – spam preventing measure)

Abstract:

This paper presents work carried out by the first author during his MSc Degree in Transport Engineering and Operations at the Transport Operations Research Group (TORG) of the University of Newcastle upon Tyne (UK). It pertains Dynamic Route Guidance Systems (DRGS) that are known to reduce travel times for vehicles equipped with in-vehicle information systems (IVIS). This seems to state the obvious: vehicles accessing additional information on traffic network conditions are better off than those without, aren’t they ? This paper will use microscopic simulation of traffic (VISSIM) to describe whether this statement is always true or whether there is a threshold beyond which the performances of the vehicles and the network as a whole worsens because too much information is available to the drivers. Two main scenarios were modelled, with or without incident on a major corridor. These were then split into 6 sub-scenarios for increases of 20% in percentages of vehicles serviced by IVIS, starting from 0 (no information). Traffic parameters used to represent network efficiency were total travel times, vehicle kilometres and queue lengths and the analysis was carried out on a network for Tyne&Wear to include parts of the A1, and A19.

Return on top

Wandsworth Road – Speed Strategy & Feasibility Study

By Joerg Tonndorf (Project Centre Limited)

Email: Joerg.Tonndorf?projectcentre.co.uk (please substitute ‘?’ with ‘@’ – spam preventing measure)

Abstract:

Wandsworth Road, situated in the London Borough of Lambeth, stretches over about two miles, with the proposed scheme incorporating eight signalised junctions and seven signalised pedestrian facilities with PUFFIN technology. The aim of the scheme was to use exclusively controller logic together with speed loops to inhibit vehicles entering the network at high speeds and prevent speeding within the corridor. The strategy is based on two functions:

1. Speeding vehicles approaching selected junctions are detected and force the controller to serve the side road, extend green on the side road, extend red following the green man or extend the all red stage.

2. Off – Peak coordination to restrict platoon speeds to an optimum speed of < 30mph

3. Overnight, all of the junctions will revert to quiescent all red (i.e. both, vehicles and pedestrians) in the absence of demand.

VISSIM was first applied to assess the impact of installing new signalised junctions and PUFFIN crossings, particularly, on bus journey times, and secondly to evaluate the potential success of the strategy, a methodology which is unique within the UK.

Vehicle Actuated Programming (VAP) allowed implementing VA functionality at all junctions during nighttimes, operating the speed strategy and emulating PUFFIN technology for pedestrian facilities. Besides, speeding vehicles and motorbikes were created and added to the vehicle compositions to represent on street conditions.

The new COM interface has been exploited to develop a variety of tools in VB to improve and accelerate the calibration process and data analysis for this project, including a saturation flow measurement tool, data output tool, strategy switch point iteration and a presentation tool for simultaneous runs.

The project is one of the first VISSIM models that have been built in accordance with the TfL modelling guidelines, revising additional standards for calibration and validation.

Return on top

Appraisal of Adaptive Public Transport Priority Measures for the Light Rail Transit Systems using Microsimulation

By Frank Dreher (Mott MacDonald) and Sonal Ahuja (Mott MacDonald)

Email: frank.dreher?mottmac.com, sonal.ahuja?mottmac.com (please substitute ‘?’ with ‘@’ – spam preventing measure)

Abstract:

This paper provides an insight into the challenges that concern, designing and modelling a modern urban light rail transit systems which include addressing multiple objectives of various demand segments – ‘often conflicting’. The design of integrated transport systems is a muti-dimensional problem where users often compete for resources and priorities. For modern light rapid transit systems the interaction of trams with various users for example, pedestrians, buses and emergency vehicles leads to a multiple objective problem where there can be multiple optimal solutions for designing an efficient system. An effective and successful LRT system is one where public transport gets (best possible) priority while satisfying the objectives of other road users as well.

This paper proposes a methodology to appraise design solutions for providing priority to public transport systems such as trams and buses, addressing the problem of pedestrian safety while minimising the delays for traffic. An adaptive signal control that provides public transport priority through a highly congested corridor is proposed and modelled using VISSIM. Finally a case study of the Midland Metro Birmingham City Centre Extension is given where for which the adaptive signal control systems were developed and are currently under consultation.

In this study signal logic was developed which not only provided tram priority, but also reduced the delays to other road users as well. Two independent signal controllers with tram detection were designed, which use the same detector location for different indicators. In addition four different signalling strategies were tested and compared by their benefits.

The VISSIM microsimulation software provided the platform for evaluation of designs as it contained the necessary complexity, the outputs for benefit estimation and the visual interface which highlighted the problems in different scenarios. It was therefore an effective tool to develop improvements to the signal control.

The study concludes that without the flexible microsimulation model it would have been impossible to design an effective integrated light rail transit priority system. The highly visual VISSIM model plays a key role in the planning process. The adaptive signal control algorithm is an innovative feature of the system that provides public transport priority while trading off between conflicting objectives of various road users. The high transparency of the microsimulation model makes the complicated LRT design process easier to appraise and understand so that ineffective solutions can be eliminated. It has helped in fighting litigation and objections to the scheme, effectively saving time and money. It provides greater confidence to the client, the designer and public in the scheme and helps develop a solution that WILL WORK.

Return on top

Return to UGM5

Return to Homepage