Finite Set Model Predictive Current Control for Variable Flux Memory Machine with a Novel ThreeLayer Optimization Strategy

Variable flux memory machine (VFMM) with changeable permanent magnet (PM) flux is recognized as a promising candidate in wide-speed-range applications. Because the electromagnetic parameters of VFMMs normally change with working conditions, the design of current controller is very demanding in terms of robustness. For this reason, an improved finite-set model predictive current control (FSMPCC) is proposed in this article. On one side, a simple measuring procedure is designed to obtain the susceptible parameters, including PM flux and dq-axis inductances. The obtained data are stored in look-up tables (LUTs) to online update the parameters involved in prediction model so that the robustness of FS-MPCC can be significantly enhanced. On the other side, in order to improve the steady-state performance, the original control set is extended by a geometric modulation method. A novel three-layer optimization strategy is then proposed to select the optimal choice from the extended control set, effectively reducing the computational burden. In addition, an improved mathematical model is presented to deal with the induced voltage caused by the rapidly varied PM flux, which is a unique issue existed in VFMM drives. Finally, the experimental results based on a prototype test rig validate the proposed FS-MPCC.