<p>A novel finite control set model predictive control (FCS-MPC) based on feedforward compensation and switching sequence optimization is proposed in this paper. The proposed method optimizes the voltage vector sets by selecting vectors with minimum switch times, thereby reducing the number of candidate states from eight to four. Consequently, four cost-function evaluations are required in each control period. Combined with load power observation-based feedforward compensation, this method integrates feedforward error compensation into a restructured cost function to mitigate load disturbances on the DC-link voltage in real time. The proposed method achieves optimization in computational efficiency with similar steady state performance, compared to conventional FCS-MPC. Furthermore, the proposed MPC method with feedforward compensation is able to improve the dynamic response under load transients, offering an effective solution for optimizing marine electric propulsion systems. Comparative experimental results validate the feasibility and performance of the proposed strategy.</p>