For a considerable period, the expressway will always be a mixture driving of conventional human-driven cars and autonomous cars due to the advancements in intelligent networking and autonomous driving technology. Based on the Gipps safety distance model, this study considers the time delay of L2 and L3 autonomous cars, and distinguishes between human-driven cars and autonomous cars. The STCA lane-changing model is enhanced and a hybrid traffic flow cellular automata model is developed based on their respective features. Then, further exploration and analysis are conducted to examine fundamental diagrams, space-time diagrams, congestion ratio (CR), and safety evaluation. The research indicates that self-driving vehicles can increase road capacity, alleviate congestion, and enhance road safety. Moreover, L3 autonomous cars have better effect than L2 autonomous cars on improving road traffic. This paper will provide a reference for studying the driving behavior of autonomous cars and the traffic flow characteristics associated with various intelligence levels of autonomous cars.