Collision-Cones-Based Nonlinear-Controller for Obstacle-Avoidance with Improved Control Gains

Recent studies on autonomous vehicles have shown that collision avoidance control strategies, based on collision cones techniques, present favorable results when ensuring driving safety in controlled scenarios. In this paper, authors present a collision avoidance control method, based on the strategy of collision cones, in which the control gains of the nonlinear Lyapunov candidate functions are dynamically determined according to the ‘time-till-collision-occurs’, ensuring the avoiding maneuver to converge in a specific laps time. Moreover, the nonlinear vehicle plant model, introduced making use of a rotational transformation from inertial to non-inertial reference frame, considers the non-negligible gyroscopic term. Quality and performances of both braking and steering avoidance maneuvers are tested by using a vehicle dynamics simulator, CarSim<sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">®</sup>, with an extended ADAS sensors package, to validate the proposed control approach on a full-scale and unknown mathematical model, even in presence of unmodelled dynamics.