SAMU (Simulation of multi-Agents and Urban Modelisation) : The GOALSim (GOdara-Arnaud- Lassarre Simulation) Model for Urban Complex Systems

Agent based and cellular automata models have been widely used in an efficient as well as effective way in context of studying granular traffic, but rare are those which consider the combined effect and interactions of pedestrians and vehicles in the urban network. So from this poit of view we are proposing here a whole new approach of studying urban complexities. A new GOALSim model is proposed which presents straightforward extensions of the cellular automata approach to traffic flow and a more probabilistic model of pedestrian movement using two new concepts of; attraction field and punctuated equilibrium effect so that GOALSim includes 1) two-way traffic 2) turning movements with traffic signals 3) pedsestrian’s intelligent involvement 4) interactions 5) risk indicators/exposure etc. using a minimal set of rules. We then analysed these new statargies using SAMU as the simulation platform. SAMU has been specifically designed for GOALSim model using ABM and CA approach in NetLogo. A few features of SAMU are also explained in this paper, which shows its high potential in the use of some practical real world problems. Some sample simulations are shown to verify our model and to some extent compare it with real world situation. Simulations showed quite promising results. The simulations at varying vehicular/pedestrian densities, and other parameters showed quite interesting results which clearly indicates that how these aforementioned features can significantly affect the vehicular/pedestrian traffic and