Design synthesis of NLMPC-based tracking controller for autonomous vehicles with active aerodynamic control

Abstract

The past three decades have witnessed extensive studies on tracking-control for autonomous vehicles (AVs). However, there is a lack of studies on effective design methods in this field. To tackle this problem, this thesis proposes a design synthesis method which is featured a design framework with two layers: at the upper layer, a particle swarm optimization algorithm is used to find optimal solutions with desired trajectory-tracking performance; at the lower layer, a comprehensively coupled dynamic analysis is conducted among the three subsystems, including a nonlinear vehicle model with active aerodynamic control for mechanical vehicle representation, a motion-planning module with given perception data, and a tracking controller based on non-linear model predictive control (NLMPC) for direction and lateral stability control. The design optimization demonstrates that the proposed method can effectively determine the desired design variables to achieve optimal trajectory-tracking performance. The insightful findings from this study will provide valuable guidelines for designing autonomous vehicles.