FPGA - based accelerator for model predictive control of induction motor drive

The model predictive control is a set of algorithms which uses the dynamical model of controlled system to predict the future systems responses. Nowadays, the nonlinear finite control set model predictive control (FCS-MPC) is a popular technique for control of induction motor drives. The FCS-MPC enables accurate and high-dynamic control of the induction motor currents, rotational speed, torque etc. with respect to system constraints (e.g. urrent/voltage limitations). The main drawbacks of FCS-MPC algorithms are their computational requirements which reduce FCS-MPC prediction horizon and its applications. Our paper is focused on the FPGA accelerator of FCS-MPC of induction motor extended by the three phase inverter model with dead-times and voltage drops. These inverter nonlinear effects play important role especially in nonnominal drive conditions (e.g. low speeds)