This paper proposes a mixed traffic flow model based on the cellular automata to study autonomous vehicle’s impacts on the characteristics of traffic flow on a three-lane segment. The analysis to the simulation results showed that when the traffic density reaches a certain value, the penetration rate of autonomous vehicles has significant impact on the speed of the mixed traffic flow. However, the spatial speed variance, braking rate and lane changing rate all develop in the same way with the increase of the penetration rate, which rise firstly and then falls. Besides, the spatial speed variance of the mixed traffic flow is generally larger than that of the pure regular vehicle traffic flow, which indicates that autonomous vehicles amplify the speed dispersion. KEYWORDS: Cellular automata; Microscopic traffic simulation; Autonomous vehicles, Mixed traffic; Penetration rate

CICTP