Fuzzy Cellular Automata Model for Signalized Intersections

At signalized intersections, the decision-making process of each individual driver is a very complex process that involves many factors. In this article, a fuzzy cellular automata (FCA) model, which incorporates traditional cellular automata (CA) and fuzzy logic (FL), is developed to simulate the decision-making process and estimate the effect of driving behavior on traffic performance. Different from existing models and applications, the proposed FCA model utilizes fuzzy interface systems (FISs) and membership functions to simulate the cognition system of individual drivers. Four FISs are defined for each decision-making process: car-following, lane-changing, amber-running, and right-turn filtering. A field observation study is conducted to calibrate membership functions of input factors, model parameters, and to validate the proposed FCA model. Simulation experiments of a two-lane system show that the proposed FCA model is able to replicate decision-making processes and estimate the effect on overall traffic performance.