Algorithm developed to estimate the future stateof a dynamic systemin an optimal way, depending on the delay of the samples admitted minimizing the noise.
Published in Chapter:
Maximum Likelihood-Based Fuzzy Adaptive Kalman Filter Applied to State Estimation of Permanent Magnet Synchronous Motors
Miriam M. Serrepe Ranno (Adaptive Systems and Signal Processing Laboratory, Federal University of Maranhão, São Luís, Brazil), Francisco das Chagas de Souza (Adaptive Systems and Signal Processing Laboratory, Federal University of Maranhão, São Luís, Brazil), and Ginalber L. O. Serra (Instituto Federal de Educação, Ciências e Tecnologia do Maranhão (IFMA), Brazil)
Copyright: © 2020
|Pages: 28
DOI: 10.4018/978-1-7998-2718-4.ch002
Abstract
In this chapter, a novel fuzzy adaptive Kalman filter for state estimation of a permanent magnet synchronous motor is proposed. The fuzzy set theory is used as a tool to perform on-line modification of the covariance matrices, adjusting the EKF and UKF parameters according to estimation reliability of the currents in the two windings of the rotor, position, and velocity for a two-phase permanent magnet synchronous motor. Also, the methodology uses the maximum likelihood technique, where the difference between the theoretical covariance and the measured covariance is defined as an approximation considering the average of a moving estimation window. This difference is performed continually and used to dynamically update the covariance matrices, aiming to obtain an efficient estimation. The membership functions are optimized to adjust the covariance matrices so that the error variation is minimal. Simulation results illustrate the efficiency and applicability of the proposed methodology.