Autonomous driving control strategies for multi-trailer articulated heavy vehicles with active safety system

Abstract

This study aims to develop automated driving strategies and integrated active control system for multi-trailer articulated heavy vehicles (MTAHVs) to enhance road transport efficiency, directional performance, and safety. To this end, a MTAHV with the configuration of A-train double was selected to be the subject vehicle, and the required vehicle models were generated. The corresponding nonlinear TruckSim model was employed as the virtual for co-simulations. The original contributions of the thesis in autonomous driving control of MTAHVs include: 1) a lateral preview driver model for MTAHVs was developed using the optimal preview control method; 2) a longitudinal motion-planning and control strategy using fuzzy sets was also devised; 3) an integrated control system was designed for coordinating autonomous driving and active trailer and dolly steering (ATDS) using a model predictive control (MPC) technique; and 4) a model-based predictive motion planning method was developed using the Frenet-Serret frame. The proposed lateral preview driver model may operate in two modes according to varied forward speed: i) in high-speed operations, the lateral stability is prioritized, and the high-speed and stability-oriented mode is activated; ii) while in low-speed curved path negotiations, the path-following off-tracking performance is emphasized, and the low-speed path-following mode is activated. It also takes benefits of the vehicle units’ body-fixed reference frames for lateral deviation calculations to mimic the driver’s local perception of vehicle position and reference path. If the so-called driver neuromuscular delay is set to zero, the driver model may perform as an autonomous human-like controller for vehicle lateral motion control. The devised longitudinal motion planner considers the road curvature over a preview horizon to regulate vehicle forward speed. It is featured with the predictive and compensatory throttle/brake actuations to assure all the vehicle units’ lateral stability. The MPC-based control method integrates the ATDS into the automated tractor steering and speed control, while the ATDS is activated to operate in either high-speed or low-speed mode, thereby improving the directional performance. The developed trajectory planner benefits from a model-based predictive approach to customize the generated trajectory to enhance the lateral stability in high-speed evasive maneuvers. The innovative findings of this dissertation will contribute to the advancement and development of autonomous driving control for MTAHVs.