Journal of Maps ISSN: (Print) 1744-5647 (Online) Journal homepage: http://www.tandfonline.com/loi/tjom20 Visualisation of Traffic Behaviour Using Computer Simulation Models Joerg M. Tonndorf & Vladimir Vorotovic To cite this article: Joerg M. Tonndorf & Vladimir Vorotovic (2007) Visualisation of Traffic Behaviour Using Computer Simulation Models, Journal of Maps, 3:1, 135-148, DOI: 10.1080/ jom.2007.9710834 To link to this article: http://dx.doi.org/10.1080/jom.2007.9710834 Copyright Taylor and Francis Group, LLC Published online: 23 Jan 2012. Submit your article to this journal Article views: 161 View related articles Full Terms & Conditions of access and use can be found at http://www.tandfonline.com/action/journalinformation?journalcode=tjom20 Download by: [46.3.195.167] Date: 17 November 2017, At: 18:19
Journal of Maps, 2007, 135-148 Visualisation of Traffic Behaviour Using Computer Simulation Models JOERG M. TONNDORF and VLADIMIR VOROTOVIC Project Centre, Saffron Court, 14b St. Cross Street, London, EC1N 8XA, UK; joerg.tonndorf@ptv.de (Received 4 th August 2006; Accepted 29 th December 2006) Abstract: Micro simulation models have paved their way into traffic engineering as a valuable tool for today s analysis and optimisation of complex traffic networks. They have proven to be a powerful tool, not only for the innovative analysis of schemes, but also to facilitate the visualisation of design elements and proposals. Microsimulation allows interested parties to actually see what is happening rather than assess the output data of pure numbers. The circulation of traffic in a network can be presented in an abstract two dimensional environment, with underlying maps and design drawings displaying the scheme. For more sophisticated presentations, static 3D models, such as signal heads, bus shelters and dynamic 3D models for vehicles can be included, offering a more realistic visualisation. In order to show streetscape elements and traffic engineering features in one model, it is also possible to produce a full 3D design model for a scheme, including buildings, street furniture and various vehicle types such as taxis, double-decker buses, cyclists and even pedestrians. The development of the micro simulation networks is based on accurate maps, drawings or aerial photographs. Visualisation and alternative output options can then be used to create two or even three dimensional illustrations of a scheme. Two project examples, being Exhibition Road and Borough High Street in London, demonstrate the application of micro simulation in traffic engineering. ISSN 1744-5647 http://www.journalofmaps.com 135
1. Introduction Traffic analysis models have come a long way in recent years, with computer simulation models providing the functionality to visualise the actual traffic behaviour rather than producing tables and lists of numeric output, which can be difficult to comprehend. Visualisation of traffic schemes is very helpful for presentations at public consultation or committee level, but it also assists the engineer in understanding or investigating the problem of an existing situation or design change. Streetscape and regeneration design can also be viewed in combination with traffic flow to encapsulate the whole design concept in one model, including public transport and pedestrian movement. One of the traffic computer simulation tools, generally known as microscopic simulation or simply micro simulation, contrasts with strategic macroscopic traffic models. The functionality and design options described in this paper are based on projects developed in VISSIM. However, similar tools such as PARAMICS or AIMSUN provide comparable features. The methods and functionality applied in micro simulation will be demonstrated in two London, UK regeneration schemes, being Borough High Street and Exhibition Road. 2. Available Functionality Microscopic simulation is a behaviour-based multi purpose traffic model with the ability to model each individual vehicle by specification, i.e. cars, buses, taxis, cyclists, trams and also pedestrians within a road network. This enables a realistic representation of actual vehicle behaviour such as lane changing, overtaking, parking / stopping and exit blocking (Wojcik et al 2004). The traffic flow model in VISSIM is a discrete, stochastic, time step based model with driver-vehicle units allocated as single entities. The movement of vehicles along the road is based on a psycho-physical following model. Lateral movement is controlled by a rule based algorithm. The driving model is derived from the continued work of Wiedemann (1974, 1991). The response of a vehicle-to-traffic situation depends on several variables such as speed difference, distance and driver dependent behaviour, which are adjustable parameters. These model parameters used for calibrating VISSIM micro simulation models are further explained in Fellendorf et al (2001). 136
Technical models are normally developed in two dimensional networks. However, for graphical presentation 3D objects and elements can be included in the model in order to provide a full three dimensional environment. 3. Methods The special movement of vehicles and pedestrians is controlled by a network model consisting of links and connectors. Various link types can be defined to model different road types e.g. urban streets, motorways or various levels of accessibility, such as bus only links, which are closed for general traffic. Static routes or dynamic routes based on origin - destination information, tell the vehicles which links and connectors to use and consequently, where and how to move around the network spatially. For controlling the temporal movement, a number of network elements, e.g. signals, give way lines, stop signs and speed decision points are placed on the network links. The advantage of micro simulation is that traffic input into the network is typically only defined for access points. At these access points, traffic can be generated at random within set volume constraints, thus creating realistic scenarios of variations in traffic arrival patterns within the network. The network building, including the design elements outlined above, is usually done on a scaled background. This can be an existing or proposed design drawing (e.g. Figure 1 shows the signal layout design for Borough High Street) or an aerial photograph (e.g. Figure 2 provides the Exhibition Road Network Overview). Lately, access to Google Earth (2006) has also helped to develop networks. The location and length of links and connectors is defined on-screen, making it easy to build the network. For presentation purposes it is possible to turn the visualisation of the model network off, thus having vehicles and pedestrians moving across the background, giving a realistic presentation of the scheme (e.g. Figure 3, Exhibition Road, simulation of a signal junction). 137
Figure 1 Borough High Street - signal layout design background. Figure 2 Exhibition Road - network overview. 138
Figure 3 Exhibition Road - simulation of a signal junction. Figure 4 Exhibition Road - demonstration of pedestrians - vehicle give way behaviour. 139
The modelling functionality allows for detailed and realistic modelling of conflict points, such as pedestrians crossing the road, by using give way elements (e.g. Figure 4 Demonstration of pedestrian behaviour, Exhibition Road), called priority rules, and accurately fine tuning headways and speed for each of these elements. Pedestrians, just like in reality, walk along links, for example representing the pavement, until they reach a crossing point where they assess the gaps between vehicles. If a gap is long enough, pedestrians follow their path on the defined link and cross the road. However, the software routine does not differentiate between give way lines or traffic signals and vehicles following each other and as a possible process, slowing down if the vehicle in front slows down. All elements that make a vehicle stop or slow down are actually caused by invisible vehicles temporary placed on the link or connector by the software algorithm. The modelling elements and functionality provided can also be used to add visualisation features that are not directly offered as an option in the software. For example, in order to model passengers waiting at a bus stop and getting on the bus once it has arrived, it is possible to use give way elements connected to an abstract signal controller. The signal controller can refer to detector recording, such as the bus arriving at the stop, and turn green without that signal actually being displayed in the network. This allows for controlling the give way element and releases the passengers once the bus has arrived. With the pedestrian link, representing the waiting area on the pavement, ending at the bus stop, passengers who have been waiting start walking towards the bus and disappear. However, in a 3D video, it looks as if passengers are getting on the bus. Similar methods have been used to make vehicles appear from a car ferry. In both cases the routine can also control the capacity by controlling the number of passengers or cars entering or leaving. 3D elements can be added as moving objects, such as cars or buses and also static objects for buildings and street furniture placed on and around the model network. There are several options to develop these 3D objects: importing objects from 3D Studio or AUTOCAD, which have either been developed with these software tools or have been purchased on the internet, where numerous 3D objects can be found. It is also possible to use the basic geometric elements available in a specific 3D modelling tool and placing pictures and patterns on boxes or cylinders developed with the 3D tool. Finally, some 3D elements are provided in online libraries free of charge. 140
4. Borough High Street Borough High Street is a Transport for London street pilot project and aims at improving the streetscape and pedestrian environment in the vicinity of London Bridge Station (Figure 5). Modelling was undertaken in order to assess the impact of the scheme on traffic, buses and also pedestrian movements (Figure 6 Vehicle and pedestrian movements, Borough High Street). Borough High Street and the surrounding road network is very complex in terms of traffic modelling; junctions are very close to each other, the number of traffic lanes varies between junctions, blocking back and exit blocking is a constant issue during peak times, and illegal parking and stopping affects the road capacity. In order to simulate on-street driving behaviour, different vehicle types, including motorbikes, cycles and articulated buses are defined (Figures 7 and 8 Interaction of various vehicle types, Borough High Street) Approximately 20 different link types were created, replicating various levels of driver aggressiveness or certain bus stop types, such as on-street stops, where only cars are able to overtake the dwelling bus. Give way elements control yellow boxes and prevent exit blocking by forcing vehicles that are stuck in traffic queues to leave small gaps for side road traffic. Interactive pedestrian movement, which by default is modelled as not interactive, was calibrated by changing the normal vehicle following parameters. Up to 20,000 pedestrians are travelling through the network in the peak hour, demonstrating the capability of the software. The maps provided show the links and connectors placed on the aerial photography with the colour coding showing the various link types applied. The inclusion of 3D elements is limited, but still provides a good representation of the area and, especially, the interaction of different road users and pedestrians. A separate map outlines the alternative display of aggregated values instead of single vehicles (Figure 9 Aggregated display of average speed, Borough High Street). 141
Figure 5 Borough High Street - network overview. Figure 6 Borough High Street - vehicle and pedestrian movements. 142
Figure 7 Borough High Street - interaction of various vehicle types. Figure 8 Borough High Street - interaction of various vehicle types. 143
Figure 9 Borough High Street - aggregated display of average speed. 5. Exhibition Road The Exhibition Road Project, carried out on behalf of the London Borough of Kensington and Chelsea, aims at improving the streetscape and pedestrian facilities by providing a shared surface and direct pedestrian crossings. The base model was developed using a combination of aerial photographs and signal layout drawings (Figure 10 Signal layout design background, Exhibition Road). For the proposal, a background drawing of Exhibition Road, provided by the streetscape architects, was converted into a readable VISSIM format (Figure 11 Aerial Photography, Exhibition Road). The core study area, being Exhibition Road, was then rebuilt, using 3D Modeller. For some buildings, a basic block is created and an image attached to the block. The image is based on photographs. However, photographs normally need to be amended, using photo editing software. Viewing angels need to be changed and certain objects, like trees close to the building, erased. For 144
more sophisticated buildings, CAD wire drawings served as a basis or more complex elements provided in the 3D modeller are combined, for example, to allow for creating balconies in front of old Victorian buildings (Figure 12 3D Modelling of Buildings, Exhibition Road). Finally, 3D objects of London landmarks, provided in 3D Studio Max format, were purchased on the internet and converted into VISSIM 3D models (Figure 13 Imported 3D Studio MAX Building, Exhibition Road). Street furniture, such as signal poles and lamp columns, which were designed similar to the ones chosen for implementation, were also added to the model, creating a very realistic display of the proposal. A high profile 3D video was recorded, combining traffic engineering and urban design elements in one model, and shown to stakeholders and at public consultation meetings. 6. Conclusions and Outlook The functionality and options available in micro simulation has made it possible to assess traffic schemes in more detail, but also to visualise existing problems and potential mitigation and improvement options. The level of detail applied to simulation networks can vary from basic two dimensional presentations to full three dimensional environments by including streetscape and architectural design elements in the traffic model. By applying some of the functions in a clever way, presentations can look even more realistic. In addition, the application of more detailed graphics such as pavement material patterns for links, 3D elements with different stages (e.g. changing displays for signals) and the export of traffic and pedestrian movements into 3D Studio will create models and 3D videos of an even higher quality. While the movement of pedestrians is already available, it is still limited to certain paths like the footway, represented by links and connectors. However, new software developments will soon allow for spatial movements too. 145
Figure 10 Exhibition Road - signal layout design background. Figure 11 Exhibition Road - birds-eye view on aerial photography. 146
Figure 12 Exhibition Road - detailed 3D modelling of buildings. Figure 13 Exhibition Road - imported 3D Studio MAX building. 147
Software Maps and visualisations produced for this article are based on the VISSIM micro simulation software. Design drawings were created in AUTOCAD, with Adobe Photoshop used to enhance and amend drawings or aerial photographs. Finally, 3D objects were created in AUTOCAD and 3D Studio Max before being exported to VISSIM. References AIMSUN (2006) AIMSUN NG - The Integrated Traffic Environment [online]. Available from: http://www.aimsun.com [Accessed: 6th June 2006]. FELLENDORF, M. and VORTISCH, P. (2001) Validation of the Microscopic Traffic Flow Model - VISSIM in Different Real-World Situations, Transport Research Laboratory, Washington D.C., USA. GOOGLE (2006) Google Earth [online]. Available from: http://earth.google.com [Accessed: 6th June 2006]. PARAMICS (2006) Paramics Microsimulation [online]. Available from: http://www.paramics.co.uk [Accessed: 6th June 2006]. WIEDEMANN, R. (1974) Simulation des Straßenverkehrsflusses. Schriftenreihe des Instituts fr Verkehrswesen der Universität Karlsruhe, Heft 8, Karlsruhe, Germany. WIEDEMANN, R. (1991) Modelling of RTI-Elements on multi-lane roads, Advanced Telematics in Road Transport edited by the Commission of the European Community, DG XIII, Brussels, Belgium. WOJCIK, S. and TONNDORF, J. (2004) Closing the Gap - Modern Modelling Software combined with advanced visualisation techniques can lessen the margin between model and reality, Traffic Technology International, UK. VISSIM (2006) ptv simulation - VISSIM [online]. Available from: http://www.vissim.de [Accessed: 6th June 2006]. 148