Model Predictive Control of a Car Active Seat Suspension System

The paper presents design of a high-level control strategy for an active car seat suspension system based on model predictive control (MPC). The MPC formulation aims to improve ride comfort by minimizing vertical seat acceleration amplitudes, while explicitly enforcing physical constraints on the suspension stroke and actuator force. Simulation results gained through a linear half-car model demonstrate that the proposed MPC strategy significantly outperforms a baseline linear quadratic regulator (LQR), primarily due to its ability to directly account for the hardware constraints. Robustness of the proposed approach is illustrated through control system evaluation based on a more realistic and complex nonlinear seat mechanism model, including actuator dynamics and friction. When combined with a proper low-level controller, the overall system achieves performance close to that of the idealized linear system, highlighting the potential for real-world application.