Feedforward predictive control for active suspension based on tire-enveloping road-excitation reconstruction

Abstract

While preview-based active suspension systems exploit road preview information for disturbance compensation, most controllers neglect tire-enveloping effects, causing distorted suspension excitation and diminished control accuracy. To address this issue, a feedforward predictive control (FF–MPC) based on reconstructed road excitation is proposed. A dual-cam geometric tire model reconstructs the road in the preview window in real time, capturing tire-enveloping effects within onboard timing constraints. Based on the reconstructed excitation, an FF–MPC controller is designed, where inverse-dynamics-based feedforward control cancels dominant road disturbances, and MPC compensates residual errors, handles uncertainties, and enforces actuator constraints by treating feedforward force and road excitation as known disturbances. Simulation and driver-in-the-loop (DIL) tests show that, compared with MPC and passive suspension, the proposed method yields larger reductions in vertical, pitch, and roll accelerations across speeds and conditions, while meeting actuator and real-time requirements.