Decentralized fast-MPC path planner for automated vehicles

摘要

Safe path-planning in an environment subject to the presence of other vehicles or obstacles in general is a crucial task for autonomous vehicle development. A complete control scheme is designed in order to test a decentralized fast MPC path planner algorithm. The proposed control scheme is composed by a receding replanner based on a fast numerical implementation of a nonlinear optimization problem and PIDs as trajectory tracker. A detailed description of the control problem formulation is provided as well as the mathematical models of vehicle considered. Numerical simulations of the control are discussed and finally conclusions are drawn.