Multiobjective Fuzzy Predictive Torque Control of an induction machine fed by a 3L-NPC inverter

Abstract

Recently the use of Model Predictive Control (MPC) in electrical motor drives has been reported both theoretically and experimentally. Predictive Torque Control (PTC) has been developed to control induction motor drives, allowing high-performance and fast dynamics. However, the optimization used in PTC is based on a single cost function minimization, where control objectives are merged by using weighting factors. The adjustment of these factors is achieved through a nontrivial process and they are heavily dependent on the system parameters and user requirements, being a complex task in systems where two or three weighting factors must be adjusted. To avoid the well-known time-consuming simulations and branch and bound search process, a Multiobjective Fuzzy Predictive Torque Control (FPTC) is presented for a Three-Level Neutral-Point-Clamped voltage source inverter (3L-NPC). The proposed strategy changes the single cost function with a Multiobjective Optimization based on Fuzzy-Decision-Making. Simulation results are presented to illustrate the behavior of the motor drive under steady-state and dynamic conditions.