Model predictive current control based on hybrid control set for permanent magnet synchronous motor drives

Model predictive current control (MPCC) has been widely recognized as a high-performance scheme for permanent magnet synchronous motor drives because of its simple control structure. This article proposes a novel multivector MPCC (MMPCC) method based on a hybrid control set, which includes both original basic voltage vectors (VVs) and synthesized VVs. The two active VVs and one null VV are considered as an initial control set, and the exact combination of them depends on their corresponding durations, which are calculated by stator current slopes in dq$dq$ frame with the use of the prediction formula. Contrary to the traditional cost function, an alternative hybrid control set can achieve superior voltage precision. This is due to the involved cost function that minimizes the maximum of prediction current errors, which typically occur at non-integer sample instants during the control period. The performance of the proposed MMPCC method has been experimented to confirm its effectiveness, compared with the conventional MPCC (CMPCC) and duty-cycle MPCC (DMPCC). The results indicated an anticipated enhancement in both dynamic and steady-state performance at a low rotor speed, alongside a significant robustness against parameter mismatch. The proposed MPCC method focuses on a hybrid control set, combining with deadbeat control method to solve the logic relationship of different voltage vector combinations and their durations. The cost function is based on the non-integral time prediction. The proposed MPCC method can maintain stability within a certain range of parameter mismatches.image