Model predictive control (MPC) has gained increasingly popularities in the field of power electronics and electrical drive systems. In MPC, the control objectives are evaluated in the cost function within a certain number of prediction horizon. Compared with conventional PI-based methods, MPC shows the merits of simple concept, straightforward implementation, fast dynamic response, flexible inclusion of constraints and ability to handle multi-objective nonlinear system. Finite control set MPC (FCS-MPC) is one of the common-used MPC schemes. In this work, the comparisons of FCS-MPC schemes and the conventional PI-based method, e.g., direct torque control (DTC) and fiel-doriented control (FOC) are carried out. It is validated that FCS-MPC methods achieves a better control performance at both the steady state and transient state.