Externalities As Objectives for DTM Solving a Dynamic Multi-objective Network Design Problem

Externalities As Objectives for DTM Solving a Dynamic Multi-objective Network Design Problem


L Wismans, University of Twente and Goudappel Coffeng; E van Berkum, University of Twente, NL; M Bliemer, University of Sydney, AU


A multi-objective optimization is carried out for a realistic case using DTM measures as decision variables and externalities as objectives. The resulting Pareto optimal set of solutions is analyzed and results like trade-offs found presented.


There is an increasing interest in the effects of traffic on externalities. Research has also shown that in many countries the costs of external effects are even larger than the costs of delay. Therefore it can no longer suffice to view a transport system as feasible or optimal if only accessibility is improved. Optimization of traffic network performance using DTM measures is a specific example of solving a network design problem. Decision variables are the specific settings of DTM measures and route choice effects are taken into account as a possible behavioral response of road users. Traditionally, a single objective function is used in such an optimization problem related to accessibility (e.g. minimizing total travel time or number of stops) and the behavioral response is modeled by solving a static user equilibrium problem. However, DTM measures have not only been identified as powerful instruments to increase network efficiency, but also to reduce externalities. As a result, in the optimization we focus not only on efficiency, but also on climate, air quality, traffic safety and noise. Additionally, traffic dynamics are important explanatory variables for these externalities and usage of a dynamic traffic assignment (DTA) model is also recommended to overcome the widely recognized limitations of static traffic assignment in particular for over-saturated traffic conditions. The evaluation of the objective functions are therefore using the output of a dynamic traffic assignment (DTA) model connected with appropriate effect models based on an extensive literature review. Using multiple objectives and a DTA model results in a dynamic multi objective network design problem (DMONDP), which is solved as a bi-level optimization problem. The solution approach used is a combination of the non-dominated sorting genetic algorithm II (NSGAII), the Streamline multi class macroscopic DTA model, the ARTEMIS emission model, RMV noise model and a risk based traffic safety model.

Solving the DMONDP results in a Pareto optimal set which provides valuable information for the decision making process, e.g. trade-offs between objectives can be determined, which would not have been available if the compensation principle would have been chosen (i.e. solving a single objective NDP with a weighted sum of all objectives). This optimization problem is NP-Hard and the evaluation of one solution is generally computationally expensive. Knowledge obtained by optimization of realistic cases can be used to attain knowledge about incorporation of externalities as an objective when optimizing traffic systems using DTM measures in practice. In our research we solved the DMONDP for a realistic network of the city Almelo (circa 100.000 inhabitants) in the Netherlands and analyzed the Pareto optimal set. In this case it is shown that the objectives efficiency, climate and air quality are mainly aligned and mainly opposed to traffic safety and noise. The objectives traffic safety and noise are neither aligned, nor opposed. However, this does not mean that there is a single solution which optimizes the three aligned objectives because still some trade-off solutions remain. The solution which optimizes air quality for example results in approximately 6% higher total travel time, in vehicle lost hours this is 29%. In contrast the solution optimizing noise, which is opposed to efficiency, results in 24% higher total travel time and 106% higher vehicle lost hours. Based on the Pareto optimal set the trade-offs are also determined. In this numerical case we found for example that we can reduce 2.5 kg CO2 emissions, accepting an increase of 1 hour of vehicle lost hours. Based on the experiences in Almelo a case study of the city of Amsterdam was set up. Solving the DMONDP turned out to be more complex, mainly due to the size of the network and the large congestion problems. In general similar trade offs can be seen as we found based on the network of Almelo.

In our paper we explain the framework used and elaborate more on the results of this optimization in terms of trade offs, solutions part of the Pareto optimal set and equity using cluster analysis. This information is useful to help traffic engineers and policy makers understand the conflicts between these objectives and possibilities to use DTM not only to improve efficiency but also noise, air quality, climate and traffic safety.


Association for European Transport