Equivalent Three-Vector-Based Model Predictive Control With Duty-Cycle Reconstruction for PMSM

The conventional model predictive control (MPC) uses a single voltage vector for every control cycle that makes the cost function minimized through the enumeration process, which will result in relatively high steady-state fluctuation and significant computational overhead. MPCs steady-state performance can be enhanced by adding duty-cycle control and generalized double-vector control, although this increases the complexity of the control algorithm. To enhance the permanent magnet synchronous motor performance while lowering algorithm complexity, an equivalent three-vector-based model predictive current control with duty-cycle reconstruction is proposed. This method eliminates the enumeration process and lowers the computational overhead in the modified voltage vector space. The steady-state performance is further enhanced using the three-phase duty-cycle reconstruction. The validity and feasibility of the proposed MPC are confirmed by comparing it with a double-vector-based MPC and three three-vector-based MPC.