A solution for urban road selection and construction problem using simulation and goal programming—Case study of the city of Isfahan Hadi Karimi Dehnavi a,1 , Mohammad Taghi Rezvan a,2 , Abdolmatin Shirmohammadli b,n , Dirk Vallée b,3 a Department of Industrial and Systems Engineering, Isfahan University of Technology, 84156-83111 Isfahan, Iran b Institute of Urban and Transport Planning, Department of Civil Engineering, RWTH Aachen University, Mies-van-der-Rohe-Strbe 1, 52074 Aachen, Nordrhein-Westfalen, Germany article info Keywords: Transportation network design problem Simulation Weighted goal programming Isfahan abstract The aim of this paper is to propose a new hybrid method of urban road(s) selection and construction by using a combination of simulation and weighted goal programming (WGP) with different environmental and economical criteria. First, different scenarios were developed based on current and proved future urban land use and on feasibility. These scenarios then were simulated by applying EMME3 software for a scheduled year. The outputs for each scenario were used as input in weighted goal programming. By using the significance of goals which determined by opinion of experts, the pair-wise comparison procedure has been done. The results of WGP determine the scenario with the highest possible utility. This integrated methodology has been applied in the city of Isfahan (Iran). The selected scenario by the proposed methodology, improves the conditions of all six criteria in comparison with current traffic network. This methodology could be applicable for large-scale transportation networks. & 2013 Elsevier Ltd. All rights reserved. 1. Introduction and literature review The urban transportation demand is changing constantly in many developing countries due to change in transportation mode, price, energy price, population and pollution, etc. The performance of the transport system affects the development and land use patterns of cities. The demand for transport systems is influenced in short- and long-term. The performance and demand produce both benefits and inconveniences for users, residents and busi- nesses (Gercek et al., 2004). To ensure that any infrastructure investment will be beneficial for the overall transport system, authorities need to predict and estimate the effects of changes. The prediction of passenger and freight demand for the distant future is a critical step for long-lived transport infrastructures planning. Assessing the consequences of a changed transport system will guide transport planners to establish practical and proper policies, which avoid an undesirable growth of transport modes. In med- ium- or long-term periods, “quantitative” and “technological” forecasting techniques, such as logistic regressions and neural networks, are more suitable tools than analyzing past data and making forecasts based on the relationship between the variables according to this data (Schafer and Victor, 2000). The problem of evaluation and selection among the urban transportation infrastructures is known as the road network design problem (RNDP). Generally there are two approaches toward network design problems (NDP). The discrete network design problem (DNDP) deals with the addition or modification of roadway segments of a street network in terms of layout and the continuous network design problem (CNDP) which deals with the optimum capacity definition of existing roadway segments. Also the mixed network design problem (MNDP) that combines both discrete and continuous approaches has been studied in some cases (Yang and Bell, 1998a). In this paper, we propose a hybrid innovative model for the discrete form. On the one hand, it is well known that a network design problem cannot be optimally solved for real-scale applica- tions by an exact multi-objective algorithm due to their complex- ity. The embedded simulation model is very often applied to the problems that are a simplification of the real case. Although a complex real-scale could be handled by a well-designed meta- heuristic multiple-criteria decision-making (MCDM) algorithm such as genetic algorithm but anyway there is no guarantee to obtain an optimal solution and moreover, no general indicators have been indicated in the literature to evaluate the effectiveness of these meta-heuristic algorithms (Cantarella and Vitetta, 2006). In addition, transportation projects are expensive so regarding constrained government expenditures; they must be carefully scrutinized for cost-effectiveness (Szeto and Lo, 2006). There is a Contents lists available at SciVerse ScienceDirect journal homepage: www.elsevier.com/locate/tranpol Transport Policy 0967-070X/$ - see front matter & 2013 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.tranpol.2013.04.003 n Corresponding author. Tel.: +49 241 80 26204, +49 176 47017100; fax: +49 241 80 22247. E-mail addresses: h.karimidehnavi@in.iut.ac.ir (H. Karimi Dehnavi), taghi_rezvan@in.iut.ac.ir (M.T. Rezvan), shirmohammadli@isb.rwth-aachen. de (A. Shirmohammadli), vallee@isb.rwth-aachen.de (D. Vallée). 1 Tel.: +98 913 2804767; fax: +98 311 3915526. 2 Tel.: +98 913 3867348; fax: +98 311 3915526. 3 Tel.: +49 241 80 25200; fax: +49 241 80 22247. Transport Policy 29 (2013) 46–53