Adaptive Cruise Control Design with Consideration of Humans' Driving Psychology

For autonomous vehicles, adaptive cruise control (ACC) system should be designed to behave similarly to humans' driving habits, in order to be acceptable. Therefore, a human-like ACC algorithm is proposed in model predictive control (MPC) framework. Not only driving safety and comfort are guaranteed in the control algorithm, but also the efficient car-following requirement (time optimal) is considered to emulate humans' driving psychology and decrease the settling time of responses. In MPC framework, the control algorithm is finally transformed to an online mixed integer nonlinear programming, which is solved by the two-loop optimization method. The simulation results show that the proposed ACC control algorithm performs similarly to humans' driving habits, and meanwhile provides safe, comfortable, and efficient car-following behavior.