Model predictive control for lane change decision assist system using hybrid system representation

This paper considers an optimal lane change path generation for intelligent automobiles using hybrid system representation. A longitudinal maneuver and a lateral maneuver of automobiles are modeled. To represent a behavior of lane changes hybrid system representations, that can combine continuous dynamics and discrete dynamics are employed. An objective function is designed to obtain a smooth lane change like a skilful drivers decision. Collision avoidance conditions are also considered. The optimal lane change path generation problem by using model predictive control is formulated as mixed integer programming. The problem is solved by using multi-parametric programming. Simulation results are shown to illustrate that this control problem generates an appropriate trajectory and input.