Implementação de um sistema de controle inteligente de posição radial aplicado em uma máquina de indução como motor-mancal do tipo bobinado dividido

Abstract

The development of technology has brought increasing challenges to the production of consumer goods with a high degree of security and control, and that they also present robustness, competitive prices and are ecologically correct. In order to meet this demand, within its competence, the field of electrical machines engineering has significantly improved the quality of these equipments, as well as the control methods, seeking an optimized operation. However, there are still major limitations in the increased use of induction machines due to the high maintenance rates of mechanical parts such as bearings, especially when access to these parts is difficult. Because of this, the research area of bearingless motors induction has been progressing, with the aim of minimizing mechanical wear and, consequently, reducing the number of maintenance. It is worth mentioning that these motors have characteristics similar to conventional induction machines, which add greater robustness and practicality, besides having a lower cost. This work describes the study and the implementation of the radial position control applied in a three-phase Bearingless Induction Machine with Divided Winding using intelligent control techniques for the optimization of the radial positioning. The machine under study consists of an induction motor of 3.7 kW, modified to operate as a motor-bearing. In this type of motor with split winding, besides the production of torque, the coil groups are also responsible for the production of radial forces for rotor positioning. This research proposes a control methodology that uses as a structure a Fuzzy-PD controller with proportional and derivative action, based on a set of control rules formulated in linguistic terms, having as main objective to present an intelligent controller for the radial positioning of the rotor. It also illustrates the comparison between the Fuzzy-PD controller and the Derivative Proportional Controller - PD for radial stabilization. For the evaluation of the controllers, performance criteria based on the error integral - IAE, ISE and ITAE - were used as a solution to evaluate the error from its transient and steady state responses. It is noteworthy that all proposed control was developed in a DSP eZdsp TMS320F28335 with floating-point architecture. As for programming, the C language was used. Regarding the module questions, the Analog / Digital converter and the PWM outputs were used. Finally, it is emphasized that the control system implemented in the DSP operates with the current and position controllers.