An Adaptive Controller for Variable Speed Electrical Drives

Abstract

A multivariable adaptive control method for processes with dynamic parameters which vary rapidly and unpredictably or immeasurably during the time is described. The control system is a model reference adaptive one. It was designed using the Popov - Landau hyperstability approach in order to synthetize the adaptation laws of the parameters of a linear model reference control system.The multivariate adaptive controller is composed of a linear model following controller (designed for the nominal set point) and a nonlinear computing block. The control vector applied-to the time varying process is the sum of the output vector of the classical regulator and of the correcting vector generated by the computing block in order to achieve uniform dynamic performance, in despite of large process parameters variations.An adaptive regulator was built in hybrid technology. It was tested on speed control of an electrical drive which momentum of inertia varies widely and rapidly. This application showed that the adaptive regulator is very attractive when the plant parameters undergo wide, unpredictable and unmeasurable changes.