Fusion of Strategic Modelling and Microsimulation: Integrated Modelling of Transport Policy Decisions at Strategic and Micro Levels
S Ahuja, T van Vuren, P Lall, B Yulainto, R Cusworth, Mott MacDonald, UK
We present our research on developing an integrated modelling approach that models the impact of policy decisions at a strategic level and, a seamless integrated microsimulation model that is used to understand the effect at detailed level.
Most planning and policy decisions have a major and micro level impact on the surrounding areas. A change in landuse pattern in an area or regeneration of derelict sites in the city centre not only calls for a micro-level investigation but also a macro-level analysis to evaluate the effect on city level or district level scale. While policy makers are concerned to check if scheme operates and gives positive returns to the economy at higher ? strategic level, traffic engineers are engaged in design and implementation of the schemes at a micro level. In order to estimate the true impact of transport scheme, both macro and micro level effects need to be modelled, quantified and analysed.
In this paper we present our research into development of an integrated modelling approach that models the impact of policy decisions at a strategic level and, a seamless integrated microsimulation model that is used to understand the effect of proposals at detailed intersection level.
The macro model was for the Swanswell Area was developed using the strategic PRISM West Midlands model developed using macro assignment VISUM modelling package and a seamless integration achieved by modelling the same at a micro level using VISSIM microsimulation software. The developed methodology is unique as it allows reutilisation of networks, highway and public transport demand data and other model components to be recycled and reutilised from macro to micro levels and vice a versa. The above results in high level of model accuracy where the policy decisions generate an area level effect at a strategic level which has a bearing at individual intersections. The combined micro decisions together can cause significant changes to travel behaviour and vehicle routing at a strategic level.
The primary objective of this modelling exercise was to quantify and test the impact of the redevelopment of the Swanswell area, and in particular, the effect on traffic patterns within the centre of the downgrading of the Coventry Ring Road between Junctions 1 and 3. For many years, whilst facilitating the movement of traffic around the centre, the Ring Road has acted as a ?concrete collar? which has both limited the development of the centre within it, and stifled regeneration of the older commercial areas adjacent to it. This phenomenon is most strikingly seen in the Swanswell area, which is typified by run-down housing areas, marginal commercial concerns and served by a congested roads network. The Coventry City Council is now addressing these problems, and has identified the Swanswell area to be regenerated, both in terms of the urban form and the associated transport linkages. One major element of the proposed master plan is the removal of the existing elevated sections of the ring road and its replacement by a boulevard more in keeping with the vision for the new Swanswell. This proposal is expected to cause a strategic and area level impact
Given the complex nature of traffic flows into and around the city centre, development of a sophisticated traffic model, which can both reflect changes in the network and demand pattern was considered to be a necessary part of the planning process.
The new West Midlands Regional multi-modal strategic model, PRISM was used as the first building block for the exercise as it contained the model network and excellent representation of travel demand pattern for the study area. PRISIM contained strategic OD movements and the networks. In order to understand the effect of the scheme at a local level, more detailed zonal and network representation was needed to be incorporated. Hence a partial network from the strategic PRISM model was generated, and details appended to reflect changes in both the land use and traffic network within the city centre at a local level. The first step in the procedure was to build a City Level model from the district/state level (PRISM) model. In the next step, using the VISUM-VISSIM integrated tool, a cordon of the city centre model area was generated and a microsimulation model of the city centre created. The city level model gave answers to the strategic impact of the models while the microsimulation model of the city centre portrayed the impacts of the policies at a micro level.
The models were unique not only because of methodology but also due to size and complexity. The final micro model contained about 100 zones and 300 intersections where about 38,000 vehicles were simulated in a typical peak hour.
We present in this paper our experiences, problems and issues associated with large microsimulation models and their links to the strategic model. The outcomes of the micro-model were utilised to change the strategic policies for the success of the scheme. We recommend guidelines that can be utilised to convert macro-micro models in a cyclic manner. Finally we present the results from the case study, which indicate that macro analysis of schemes can be utilised as a basis for detailed area analysis and micro analysis of traffic operations can be used in appraising the effect of scheme at a larger spatial level.
The case study provides insights into integrating the inputs and outputs of the VISUM strategic model with the VISSIM local model to assess the impacts on the city centre road network of strategic transport issues, and to pass back the implications of local land use and network changes. This methodology results in true fusion of strategic modelling and microsimulation techniques.
Association for European Transport