EVALUACIÓN DE CONTROLADORES PREDICTIVOS EN ACCIONAMIENTOS DE MÁQUINAS DE INDUCCIÓN USANDO AMBIENTE SIMULINK

Abstract

Control theory and computing power of microprocessors enables new control strategies for electrical machines drives. The strategy of model-based predictive control (MPC) has the advantages: good dynamic response, and the nonlinearities and system constraints can be easily included. The discrete nature of the converters facilitates the implementation of the optimization problem being able to evaluate all possible states. In this paper we assess the development of MPC for an induction machine 5 [Hp] through a three-phase inverter with two levels. A predictive model of the machine lets you know how to control variables: stator flux and electric torque. These variables are predicted for each possible switching state of the converter. Function is defined by the distance as the error between the controlled variables and the desired behavior of these, and is used as a criterion for selecting the optimal state. The switching state, which minimizes the quality function is selected, is applied during each sampling period. This control strategy is evaluated by simulations using the environment Simulink/Matlab, resulting in a different approach that allows controlling effectively the engine drives. Keywords: predictive control, electric drives, induction machine