The control strategy of human operator is naturally dynamic, stochastic and nonlinear. It also changes gradually through the time. Therefore, human driver modeling is a very difficult task The driver model must be accurate and robust enough to give valid human behaviours under varying conditions, otherwise the model will most probably fail. This means the model has to have some intelligence similar to the human driver. In this paper, human driver modeling was analyzed from an intelligent control perspective.