Design optimization of articulated vehicles with autonomous steering control

Abstract

Human driver errors cause about 94% of traffic collisions. To increase the safety of road vehicles, extensive studies have been conducted on developing autonomous driving technologies. However, little attention has been paid to exploring autonomous articulated vehicles (AVs). This thesis proposes an approach to the design synthesis of AVs with autonomous steering. A linear yaw-plane model is generated to represent the AV, and a model predictive control (MPC) based tracking controller is designed for steering control. A stochastic modeling technique is developed using Monte-Carlo method to evaluate the performance limitations of AV dynamics. For enhancing the performance of the selfsteering AV, the design synthesis is formulated as a bi-layer design optimization problem. Particle Swarm Optimization (PSO) and Differential Evolution (DE) are introduced and tested. Selected simulation results are presented and discussed, and the insightful findings may be used as guidelines for developing autonomous driving control systems of AVs.