Ahmad Farahat - Malek-e-Ashtar University of Technology
Hamed Farahat - Isfahan University of Technology
Nematollah Ghahramani - Malek-e-Ashtar University of Technology
Mohammad Danesh - Isfahan University of Technology

Recursive Least Square (RLS) method is a powerfulapproach for on-line parameter identification problems.Resetting of covariance matrix, in systems with severe andabrupt parameter variations, presents an effective solution tomeet the situation. In this paper, a method is introduced toestimate the constant load torque and inertia in induction motors,based on RLS and covariance matrix resetting. In the proposedmethod, the resetting of covariance matrix is conditional insteadof being periodic. The studied induction motor is modeled usingthe discrete electromechanical equations where, the frictioncoefficient is neglected. Simulations for various operatingconditions show the effectiveness of the presented method.

Induction Motor; Inertia; Load Ttorque; Recursive Least Square Method;