Comparison of triangular and Gaussian membership functions performance for a fuzzy PD controller for an induction motor drive.

摘要

Three phase induction motors are common in industry for converting electrical power into mechanical power. They are considered to be simple, rugged, robust, efficient, and suitable for applications in harsh environments. However, despite their popularity, their controllability remains a difficult task using conventional control methods. The control difficulty is associated with the high nonlinearity of the motor's behaviour, complexity of its analytical model and presence of interactive multivariable structures. Therefore, designing a conventional controller takes significant time and effort and requires the presence of an accurate model of the motor.;Fuzzy controllers, unlike conventional ones, can be designed without the need for an accurate model or with less information about the controlled system, which applies in the induction motor's case. Since designing a fuzzy controller is usually by trial and error, it is unclear how the controller parameters such as the shape of membership functions (MFs), their number and controller gains affect its work.;This research focuses on studying the effect of some controller parameters on Indirect Fiel- Oriented Control (IFOC) of an induction motor (IM). Those parameters include the shape and number of membership functions (MFs) in addition to the controller gains. Only triangular and Gaussian membership functions are considered for the shape. The fuzzy controller uses both the speed error and its rate of change as inputs, and the electromagnetic torque as an output in a relationship mapped by a fuzzy rule table.;In recent years, the advances in power electronics and computing fields have made it possible to use three phase induction machines in high performance applications. Also, more robust and intelligent controllers have been deployed for induction motor control. One of those intelligent controllers is the Fuzzy Controller (FC). Fuzzy controllers allow the human experience to be incorporated in the control process.