Abstract—The purpose of this paper is to develop a computational model to study vehicle flow on urban roads. The model presents a study undertaken at a two-way intersection where vehicle flow is controlled by traffic lights operating according to fixed intervals of times. This model is computationally simulated using Matlab/Simulink. Several simulations were processed and in practice the model proved to be stable and reliable. It is expected that the results will contribute to the creation of a stochastic predictive model to anticipate situations that lead to traffic jams. Index Terms—Simulation model, traffic lights, traffic simulation. I. INTRODUCTION Computer simulation means running a computer program that will reproduce an abstract model of a given system, enabling these systems to be studied to assess their behavior relevant to the variations in the study environment. According to [1], the advantage of computer simulation is the ability to answer questions without disrupting the systems investigated. This paper presents a computational model to study urban traffic, aiming to optimize traffic flow and reduce traffic jams by reducing the average time vehicles remain stationary at traffic lights. II. METHODOLOGY This paper presents an exploratory study of traffic lights and their impact on urban traffic conditions. To do so, we conducted a bibliographic study to establish parameters allowing us to develop a computational model that accurately represented reality. Vehicle flow in urban traffic and the quality of service were studied by analyzing frequency of stops, speed of operation, travel time, traffic density and the operational costs of the vehicle. The computational model was developed and simulated using Matlab/Simulink, a tool projected for this type of study. So that the model accurately represented real-life demand for the system, information regarding traffic light cycles and other aspects were taken data published in the National Traffic Department Traffic Lights Manual - [2]. Additionally, the traffic light times were calculated using the Webster method, the algorithm of which is provided in [2]. Manuscript received November 10, 2013; revised January 2, 2014. The authors are with the Universidade Presbiteriana Mackenzie, Brasil (e-mail: {takato, raunheitte, rubens.camargo, aavallim}@mackenzie.br, juliopetroni@yahoo.com.br). III. COMPUTER SIMULATION AND SYSTEMS MODELING Modeling is the development of a model to represent a given phenomenon. A model is the representation of a real or imaginary system using a language. A Matlab environment application, Simulink, was used for the simulation in this paper, given that this tool provides an interactive environment based on block diagrams, designed specifically for modeling, simulation and analysis of continuous, discrete or hybrid dynamical systems [3]. According to [4], the models are classified as either „static‟ or „dynamic‟. Dynamic models are then sub-classified as „deterministic‟ and „random‟. The random models adopted in this paper are divided into discrete and continuous change, as shown in Fig. 1. Fig. 1. Classification of simulation models. Source: [4]. IV. URBAN TRAFFIC The degree of randomness intrinsic to real systems, scaled specifically to random occurrence of events in these systems, necessarily implies proportionate application of Monte Carlo method. The use of this technique consists in the artificial generation of data by employing a random number generator (RNG) and a frequency distribution of the target variable. The RNG is a computer program capable of generating independent and uniformly distributed random values between zero and one [1]. This study uses the terminology proposed by [5]. The term transit refers to the displacement of people or vehicles and traffic is applied when it refers to the study of these displacements. The concept „traffic jam‟ is related to the concepts of road capacity and service level. Road capacity represents the maximum number of vehicles that moving along a stretch of raid over a given period of time, in a specific set of conditions of composition of traffic demand and environment; service level is a measure of the quality of service provided to the road user. This quality of service can be analyzed according to frequency of stops, speed of operation, travel time, traffic density and operational costs of the vehicle. According to [5], service level is a parameter that reflects the conditions of flow, safety and comfort on any given space that is occupied. Takato Kurihara, Luis Tadeu Raunheitte, Rubens de Camargo, Arnaldo Rabello de Aguiar Vallim Filho, and Julio Petroni A Computational Model to Study Urban Traffic Control International Journal of Computer Theory and Engineering, Vol. 6, No. 5, October 2014 382 DOI: 10.7763/IJCTE.2014.V6.894