Accuracy Improvement of Induction Motor Speed Estimation using Improvised Tuning of Extended Kalman Filter Technique

Abstract

Speed control schemes considered till-date use a sensor to measure the speed of the rotor. In many cases, it is impossible to use speed sensors for measurements, because it is either technically impossible or extremely expensive. For example, the pumps used in Oil rigs to pump out the oil have to work under the surface of the sea, sometimes at depths of 5 m and getting the speed measurement data upto the surface requires extra cables. In such cases the most vulnerable part of the drive is the sensor. Cutting down the number of sensors and measurement cable provides a major cost reduction. The work presented discusses a computationally effective recursive Extended Kalman Filter (EKF) algorithm which is specifically designed for better speed estimation of induction motor. EKF developed provides the minimum variance state estimation and tolerates Induction motor modeling and measurement errors. EKF can provide an improved estimate of the rotor speed and achieve fast convergence. A real coded Genetic algorithm is used to tune the noise matrices for improvising the EKF performance. Simulation analysis of the proposed method is provided.