Personalized Adaptive Cruise Control Based on Artificial Potential Field Model Predictive Control

Adaptive cruise control (ACC) without consideration of driver characteristics will lead to a decrease in driver’s acceptance of the system. A personalized ACC (PACC) method is proposed based on artificial potential field (APF) in this paper. Firstly, the data of diverse drivers when following leading vehicles are collected on a driving simulator platform. And the APF model is established based on drivers’ preferences for time headway, desired velocity, and maximum acceleration. Secondly, the parameters of APF model are calibrated by NGSIM dataset. Then, a personalized car following controller is designed based on model predictive control (MPC) and APF. Finally, the proposed APF based PACC method is evaluated in Matlab/Simulink-CarSim co-simulation environment. Simulation results demonstrate that the APF based PACC can simulate the driver’s behaviors in following the leading vehicle with different speeds, and adjust the time headway (TH) within the corresponding comfortable zone for the following vehicles driven by drivers with different styles.