LPV/H∞ Controller Design for Path Tracking of Autonomous Ground Vehicles Through Four-Wheel Steering and Direct Yaw-Moment Control

Peng Hang,Xinbo Chen,Fengmei Luo
DOI: https://doi.org/10.1007/s12239-019-0064-1
2019-07-09
International Journal of Automotive Technology
Abstract:<p class="a-plus-plus">This paper focuses on the path-tracking controller design for autonomous ground vehicles (AGVs) using four-wheel steering (4WS) and direct yaw-moment control (DYC) systems. In order to deal with the parametric uncertainties, a linear parameter-varying (LPV) <em class="a-plus-plus">H</em><sub class="a-plus-plus">∞</sub> controller is designed as the high-level controller to generate the front and rear wheel steering angles and external yaw moment based on linear matrix inequality (LMI) approach. The lower-level controller is designed for torque allocation between the left and right side wheels to yield the desired total longitudinal force and external yaw moment utilizing weighted least square (WLS) allocation algorithm. To test the performance of the proposed path-tracking controller, numerical simulations are carried out based on a high-fidelity and full-vehicle model constructed in CarSim. Simulation results show that the LPV/<em class="a-plus-plus">H</em><sub class="a-plus-plus">∞</sub> controller has better path-tracking performance than the fixed gain <em class="a-plus-plus">H</em><sub class="a-plus-plus">∞</sub> controller. To show the superiority of 4WS+DYC control system, the contrast simulation is performed based on LPV/<em class="a-plus-plus">H</em><sub class="a-plus-plus">∞</sub> controller. Simulation results indicate that 4WS+DYC control system has better path-tracking performance and handling stability than active front steering (AFS), AFS+DYC and 4WS control systems.</p>
engineering, mechanical,transportation science & technology
What problem does this paper attempt to address?