LPF-based adaptive observer for speed and acceleration estimations of electric motor drive applications without system model dependence

This article designs a novel adaptive speed and acceleration observer, incorporating a low-pass filter (LPF) as an auxiliary system, which is independent of the system model of the exact parameter values and load information of the motor, based on the motor drive position measurement. The contributions of this article are given as follows: first, an incorporation of LPF into the proposed observer to eliminate the high-frequency noises of the measurements, second, the collaboration of the nonlinearly structured observer gain and special coordinate transformation to make the adaptive observer design possible with the error dynamics diagonalization by the pole-zero cancellation, and third, the adaptive system as another subsystem to secure the improved robustness against the low-frequency noises of the measurement. The experimental verification highlights the practical advantages of the proposed adaptive observer based on a 500-W electric motor drive system as a prototype under various loads.