Estimation‐Driven Model Predictive Control Architecture for Active Suspension Systems Under Road Disturbances

This research work presents an optimal control architecture for a quarter‐car active suspension system, which integrates a discrete‐time Kalman filter (KF) with a model predictive control (MPC) scheme. The proposed framework aims to enhance ride comfort and vehicle stability by effectively suppressing the impacts of road disturbances, that is, road bumps and pathholes, through a feasible control input. The proposed framework is based on an estimation‐based control architecture, where a KF provides an optimal reconstruction of the unmeasurable states. The estimated states information is then fed as input to the MPC scheme, which minimizes a cost function by solving a finite‐horizon constrained quadratic optimization problem to compute an optimal control input. The stability of the proposed control architecture is thoroughly investigated via a Lyapunov‐based analysis, with the MPC cost function serving as a Lyapunov candidate function. Computer simulations are carried out in the MATLAB/Simulink environment, where the results for both real and estimated states confirm the minimum‐variance property of the KF. Furthermore, the results corresponding to the open‐loop and closed‐loop configurations demonstrate the efficiency of the closed‐loop optimal control strategy in state regulation, disturbance rejection, and control effort. In perturb scenarios, the performance parameters, such as oscillation suppression, settling time, and overshoot, are comparatively analyzed. The results demonstrate the effectiveness of the designed control framework in damping undesired oscillations, with minimum overshoot and reduced settling times, thus improving ride comfort and vehicle stability.