Dual-Objective Road Sense Construction for Steer-By-Wire System Based on Friction Compensated Adaptive Model Predictive Control

By decoupling manipulation and actuation, steer-by-wire systems represent a next-generation steering structure for autonomous vehicles. However, existing road sense constructions primarily focus on torque tracking, failing to reconcile the tactile feedback and returnability. To bridge this gap, this article proposes a novel dual-objective road sense framework that regulates switching torque and angle targets for realistic feedback and self-centering behavior. Initially, a comprehensive model of the hand wheel module is developed as the foundation of the subsequent designer and controller. Next, a modular steering feedback torque generator and a fuzzy logic-based steering returnable angle designer are formulated, respectively. Furthermore, a friction-compensated adaptive model predictive controller is introduced, where state weights are dynamically tuned via a support vector machine, enabling seamless switching between torque and angle targets within a unified scheme. Finally, hardware-in-the-loop experiments are conducted to validate the proposed framework, demonstrating its effectiveness in replicating realistic road sense and returnability, while achieving superior tracking performance across diverse driving scenarios.