Robust adaptive control for vehicle active suspension systems with uncertain dynamics

This paper proposes adaptive control for vehicle active suspensions with unknown nonlinear dynamics (e.g., nonlinear spring and piece-wise damper dynamics). An adaptive control is designed to stabilize the altitude of vehicles and to improve the ride comfort, where an augmented neural network is developed to provide the online compensation for the unknown dynamics. A novel adaptive law is proposed to estimate the NN weights and essential model parameters (e.g., mass of vehicle body, inertia for pitch motion). The parameter estimation error is derived and used as a novel leakage term superimposed on the adaptation to guarantee the error convergence. Theoretical studies are provided to address the closed-loop system performance and to compare the novel adaptive law with traditional adaptive laws. The suspension space limitations and dynamic tire loads are also analyzed. Finally, comparative simulations are included to verify the effectiveness of the proposed control.