Nonlinear predictive current controllers for Permanent Magnet Synchronous Motor

The work presented in this paper relates to nonlinear predictive current controllers applied in Permanent Magnet Synchronous Motors (PMSM) control systems. The first control scheme is the simplest one - it predicts future evolution of currents for each inverter key configuration, and selects the one that minimizes given cost function. The next control scheme uses three configurations in order to minimize current ripples. The duration of selected key configurations is calculated in order to move current vector as close as possible to the set current vector. To define the duty cycles of each inverter configuration there is applied a heuristic method. Motor parameters were calculated with FE (Finite Element) model created in Maxwell and were also verified experimentally on a test stand. A motor model was based on industrial stator and windings, and rotor with buried magnets, previously optimized, designed and constructed. Simulation of the proposed control algorithms were carried in Matlab using scripting language - without Simulink.