On the basis of the research of driving behavior, by analyzing the mechanism of driving behavior and using the cellular automaton (CA) model, we introduce the mechanism of driving behavior, which is associated with the vehicle’s own speed, the relative speed, the vehicle spacing, and the safety vehicle spacing, into the CA model (ACA). We observe the metastability, phase-separation and hysteresis phenomena when taking advantage of ACA model to conduct the microscopic traffic simulation and compare the simulation results of ACA model with the simulation results of the CA model (BCA) which only considers the vehicle’s own speed and the vehicle spacing. The results show that, in the macro level, from the maximum traffic flow, stability and the efficiency in dissolving congestions, the simulation results of ACA model exhibit a larger maximum value of flow which is also closer to the actual measured data, stronger stability and higher efficiency of dissolving jam; in the micro level, from the fluctuations of the vehicles’ speed and the vehicles’ headway-distance, the simulation results of ACA model exhibit that the vehicles do not accelerate or decelerate suddenly, the vehicles can be relatively and evenly distributed on the road. All these results illustrate that the relative speed and safety vehicle spacing have a significant impact on the driving behavior and also explain that the ACA model is in better consistence with the actual situation and has some practical significance.