Accurate vehicle velocity estimation is critical for advanced driver assistance systems (ADAS) and autonomous driving, yet faces significant challenges due to unknown road friction and tire saturation effects. To address these issues, this paper proposes a transient-aware fusion observer that incorporates the real-time tire-road friction conditions to achieve robust velocity estimation. Longitudinal and lateral tire forces are estimated by dedicated observers. These forces are used to infer the tire-road friction coefficient (TRFC), which serves as input to a LuGre tire model to capture transient friction dynamics. A nonlinear observer then fuses kinematic velocity predictions with model-based corrections, using acceleration mismatches and friction-coefficient errors as feedback to refine the estimates. By integrating the friction state into dynamic corrections, the observer maintains accuracy even during tire saturation, ensuring reliable state estimates. The asymptotic convergence of the observer is rigorously proven using Lyapunov stability theory. The performance of the proposed observer is examined through field test scenarios under varying adhesion roads. The results demonstrate that the proposed observer exhibits higher estimation accuracy compared with benchmark methods.
While rear-wheel distributed-drive vehicles offer superior handling potential, maintaining stability under extreme maneuvers remains a critical challenge. Conventional Electronic Stability Control (ESC) systems effectively manage instability but often cause undesired speed reduction. Conversely, Rear-Wheel Torque Vectoring (RWTV) maintains speed but struggles to generate sufficient external yaw moments without inducing rear-wheel saturation and drift instability. To address these limitations, this study proposes a coordinated control strategy integrating RWTV and ESC based on a proactive drifting prediction mechanism. It adopts a three-layer control architecture, including a supervisory layer for vehicle's drift risk identification, a direct yaw moment decision layer for reference yaw moment calculation, and a torque distribution layer using Model Predictive Control (MPC) for multi-objective optimization. Simulation results illustrate that the proposed strategy effectively enhances vehicle maneuverability and stability under various driving conditions, providing an effective coordination method for RWTV and ESC.
7 November 2025 in IEEE Transactions on Industrial Electronics
Recent advances in autonomous driving and electric vehicles have highlighted the critical importance of vehicle stability control systems. However, existing control strategies often struggle to maintain optimal performance in challenging driving scenarios. This article presents a novel coordinated control strategy that seamlessly integrates torque vectoring control (TVC) and differential braking control (DBC). Our approach significantly enhances vehicle stability limits while minimizing energy-consuming brake interventions. By prioritizing TVC for yaw moment generation and employing DBC as an intelligent backup system, we achieve superior vehicle dynamics control. The strategy features an advanced boundary map for precise DBC activation and implements a sophisticated multiobjective control algorithm at the torque distribution level. Through extensive testing, our system demonstrated remarkable capabilities in preventing excessive longitudinal tire slip while maintaining optimal lateral forces, marking a significant advancement in vehicle safety and handling performance. The proposed controller is experimentally validated under various road and operating conditions. Test results demonstrate that the controller can significantly improve vehicle handling stability in extreme conditions.