A sliding-mode control approach towards design of driving assistance system for lane-change maneuver in connected vehicles

Advanced driver assistance systems (ADAS) that can assist a human driver in performing complicated maneuvers are known to reduce driver stress, enhance safety, and improve road throughput. We approach the design of ADAS, specifically for lane changing in connected vehicles, from a trajectory tracking perspective. In particular, we design a controller that ensures that the dynamics of the subject vehicle (SV) tracks a reference trajectory for safe lane changing, while avoiding collisions with lead vehicles (LVs) in the current and target lanes. The dynamics of the SV is modelled using the basic kinematic bicycle model. The reference trajectory comprises: (i) the longitudinal trajectory based on the two-lane car-following framework that considers the positions of the two LVs, (ii) the heading angle trajectory based on a logistic growth-decay model which accounts for the lane width and time available for the maneuver, and (iii) the lateral trajectory obtained using the two aforementioned trajectories. A coupled multiple sliding-mode control based controller is designed for trajectory tracking such that the SV traverses along the desired path. Our theoretical results are substantiated with numerical simulations. It is seen that the derived controller tracks the desired trajectories efficiently.