Optimization of Traffic Network Design Using Nature-Inspired Algorithm: An Optimization via Simulation Approach

Optimization of Traffic Network Design Using Nature-Inspired Algorithm: An Optimization via Simulation Approach

Enrique Gabriel Baquela (Universidad Tecnológica Nacional, Argentina) and Ana Carolina Olivera (Universidad Nacional de la Patagonia Austral, Argentina & CONICET, Argentina)
DOI: 10.4018/978-1-4666-9644-0.ch010
OnDemand PDF Download:


In this chapter the application of a nature-inspired technique in conjunction with simulation models to optimize the siting of concentration nodes in a traffic network of urban area is presented. The solutions of this optimization problem involve the redesign of the network by adding nodes and arcs to the current traffic network. The problem is a sub-problem of the most general “Traffic Network Design Problem”. For this resolution a Genetic Algorithm approach was design and development. The popular Simulator of Urban Mobility (SUMO) is use as traffic simulator in order to evaluate the solutions obtained. This chapter contains the conceptual models and the results of the Optimization via Simulation technique proposed.
Chapter Preview


These days, urban expansion has become a special issue due to the impact in ecosystem, biodiversity and human activities (Bhatta, Saraswati & Bandyopadhyay, 2010; He, Tian, Shi, & Hu, 2011; Haregeweyn, Fikadu, Tsunekawa, Tsubo, & Meshesha, 2012; Poelmans & Rompaey, 2010). However, the absence of urbanization planning policies around the world has turned this expansion process chaotic and disorganized A very common phenomenon is the urban sprawl, which means that a city has spread into several suburbs connected with the commercial and economic center of the town through highways, resulting thus in vehicle dependency (He, Okada, Zhang, Shi, & Li, 2010).

When planning to expand a city, it is necessary to achieve a sustainable use of the land, to exploit the roads and highways with low traffic, and to reorganize the urbanization process in order to guarantee a minimization of the environmental impact and a reduction of traffic problems (He, Tian, Shi, & Hu, 2011).

In this context, in recent years a lot of work has focused on redefining expansion areas in urban settlements (Ferreira & Condessa, 2012), and on employing Geographical Information Systems with sensors to test the impact of urban planning policies (Stanilov & Batty, 2011; Xiao et al., 2006). If investing in new infrastructure (sensors, data recover stations, etc.) is not possible, then the simulation and optimization heuristics become the alternatives to analyze the impact of urban expansion models on the landscape (Poelmans & Rompaey, 2010). In this chapter it is proposed to use a framework based on Optimization via Simulation (OvS) which combines Genetic Algorithms (GA) and Traffic Simulation in order to evaluate the organization urban planning in the cities, taking into account its impact on the traffic network.

The objectives of this chapter are:

  • Introduce the traffic optimization problem.

  • Introduce the “Origin-Destiny Traffic Assignment Problem”.

  • Explain how to model the “Origin-Destiny Traffic Assignment Problem”.

  • Show the main aspects of Genetic Algorithms and Optimization via Simulation.

  • Introduce the traffic simulation.

  • Explain how to solve the “Origin-Destiny Traffic Assignment Problem” with Genetic Algorithms and Optimization via Simulation.

  • Test the effects of parameters selection over the solutions found by OvS.



City growth reflects a progressive dynamic that is hardly ever planned, resulting in a decrease of the general performance of its sub-systems (Xiao et al., 2006). In the case of the traffic sub-system, a population growth means more vehicles in the system, higher probability of traffic jams, longer travel times etc. (He et al., 2010). Even though this is not easy to control, it is possible to establish policies related to urbanization permits, to create industrial centers and to encourage activities in strategic zones, which will allow us to control the way the city grows and its associated traffic sub-system (Xiao et al, 2006). In this line, the traffic simulation (Li & Sun, 2012; Nagel, & Schreckenberg, 1992) revels itself as a useful tool to quantify the impact of a certain traffic network topology and the corresponding population assignment (Garcia-Nieto, Olivera & Alba, 2013).

Key Terms in this Chapter

Traffic Demand: set of all vehicles in a traffic systems, with their associated routes.

Traffic Flow: set of all vehicles going from node A to node B in the same time window.

Optimization via Simulation: optimization framework in which the evaluation of the objective function is made in a simulator.

Traffic Net: oriented graph that represents the physical structure of a traffic system.

Traffic Simulation: dynamic representation of a traffic system achieved by building a computer model and moving it through time.

Traffic Destination: sink of traffic flow in the traffic system.

Traffic Origin: source of traffic flow in the traffic system.

Traffic Jam: stop of vehicle's circulation cause by a high density of vehicles in the system or difference of speed between a set of vehicles.

Genetic Algorithm: canonical optimization method that emulate the evolution and inheritance.

Traffic System: an specific traffic net with its traffic demand and its control policies.

Complete Chapter List

Search this Book: