Estimating the parameters of an induction motor in open-loop and closed-loop operation

Abstract

The purpose of this paper is to present two methods for estimating all the parameters of induction motors operating in open-loop or in closed-loop. The parameters are obtained by solving a recursive least squares minimization problem. The estimation procedure is performed in two parts. First, the stator leakage inductance and the stator resistance are determined by using the homopolar machine model. Then, given the previously estimated homopolar parameters and by using the dynamic dq model, all the other parameters are determined. The first method is derived for open-loop operation. In the second method, the model is derived for the case where the stator currents of the motor are regulated via a linear feedback controller. Selected experimental results are used to demonstrate the feasibility and performance of the proposed methods.

A MRAC parameter identification algorithm for three-phase induction motors

“…Among these methods, most of them do not provide all parameters of the machine [2], [3]. Conversely, it had been proposed a procedure based on RLS algorithm to obtain the all electrical parameters of induction motor [4]. However, this method requires previous classical tests to obtain the initial parameters used in the closed-loop RLS identification algorithm.…”

“…However, this method is imprecise because it uses some approximations to derive the model. In order to solve this problem, several recursive algorithms to obtain the model of induction motors have been proposed in literature [2]- [4]. Most of these techniques have used Recursive Least Square (RLS) identification algorithm to obtain the model of the machine.…”

Adaptive control of the three-phase squirrel cage induction motor with load torque estimator

“…Despite of the inflexible conventional methods, the on-line parameter identification ones are more advantageous. The recursive least square identification algorithm is the most used method [2], [3]. However, in order to obtain the initial conditions of the parameters, the classical tests must be done [1].…”

On-line identification of induction motors through discrete models for sinusoidal signals

“…This fact, highly inconvenient, has caused the development of several techniques to overcome the inherent difficulties ( [3], [4], [5], [6], [7]). This In [3], a technique called Reinitialized Partial Moments is developed, but of difficult application.…”

“…In [5] and [6], the derivatives are obtained by Filtered Differentiation using low-pass digital Butterwort filters, and in [7] by state variables filters. Methods that use filters always introduce delays of difficult compensation.…”