Autonomous rear-end collision avoidance using an electric power steering system

摘要

A step towards fully autonomous vehicles can be achieved by autonomous collision avoidance. This paper presents the development and implementation of a rear-end collision avoidance system which consists of novel linear threat assessment, projected escape path planning with non-zero initial condition, reference generation and linear path control by an electric steering actuator using dynamic inversion. The linear threat assessment continuously calculates threat metrics associated with collision avoidance by either steering or braking. The escape path planner continuously estimates a path with constraints on the peak lateral acceleration and the required lateral deviation. The reference generator then estimates the states for avoidance according to the available friction. Linear state feedback control is applied for path following to avoid a collision in both straight-road scenarios and curved-road scenarios. The proposed approach is evaluated in simulations, and in practice on a C-segment Ford vehicle. A user-related assessment about the usefulness and satisfaction is included to conclude the paper.