Mining Dynamic Transition Rules of Cellular Automata in Urban Population Simulation

In recent years, cellular automata (CA) has been widely used to simulate urban system and the relevant fields. The primary issue is to explore a set of dynamic transition rules that incorporate a neighborhood effect. However, most of transition rules are defined statically by one or several equations. In this study, autoregression (AR) is introduced to establish a CA-AR model integrating CA for data mining of transition rules dynamically. The proposed CA-AR model was implemented using MapInfo Professional, MapBasic and SPSS. As a case study, CA-AR model was applied to simulate population development in Nanchong City of Sichuan Province, China. The results show that this model can achieve high accuracy and overcome some limitations of static transition rules of existing CA in simulating population development. In a word, it is concluded that the proposed model is feasible to simulate the population development of urban system.