A Steering Load Torque Reconstruction Approach Across the Full Speed Range for Steer-by-Wire System

To address road feel loss in steer-by-wire systems and overcome the limitations of unstable sensor measurements and low-speed dynamic model failures, a full-speed-range front wheel load torque reconstruction method is proposed. First, a nonlinear two-degree-of-freedom model integrated with the Magic Formula tire model is built to analyze load characteristics across speeds. Then, the speed range is divided into 9 intervals for reference load calculation, and a fuzzy switching controller is established to ensure smooth reconstruction during speed transitions. Finally, a series of simulations and experiments is conducted to validate the applicability and accuracy of the proposed method under varying speeds. In a steady static circular test, the reconstructed load torque matches the simulated torque, achieving a coefficient of determination greater than 0.9982 and the torque normalized root mean square error (NRMSE) below 2%. The steering-wheel sinusoidal input test verifies that the reconstructed load tracks the simulated load in both amplitude and phase, the amplitude differences are both within 5%, and the phase differences are both within 50 ms between the reconstructed load and simulated load. Real-world driving tests indicate that the proposed reconstruction algorithm reduces the torque NRMSE by approximately 9.2% compared to the current-based method, resulting in a value below 3%.