Consistency of Extended Kalman Filtering and Particle Filtering Techniques for the State Estimation of Brushless DC Motor

Abstract

The paper deals with different adaptive filtering algorithms (AFAs) and is implemented to the non-linear systems. As Brushless DC (BLDC) motors are non-linear in nature under the presence of noise signals like gaussian white signals, these AFAs are implemented to BLDC motor by the estimation of state and output measurements. The performance characteristics (both steady state and dynamic state characteristics) of BLDC Motor are improved by using filtering techniques like Extended Kalman Filtering (EKF) and Particle Filtering (PF). The consistency of EKF and PF methods is determined by using consistency tests, and the obtained simulation results proves that the consistency of PF technique is much better than the EKF technique. PF technique shows the better stability, strong robustness with better convergency and reliability and low sensitivity on the estimation of state and output measurements of Sensorless BLDC Motor control with non-linearities.