A high performance globally exponentially stable sensorless observer for the IPMSM: Theoretical and experimental results

TL;DR

This paper introduces a new globally exponentially stable flux observer for IPMSMs that improves transient performance by using a novel invariant manifold approach and advanced estimation techniques, validated through simulations and experiments.

Contribution

It proposes a high-performance GES flux observer for IPMSMs that overcomes previous limitations by employing a virtual invariant manifold and Kreisselmeier's regression extension estimator.

Findings

Enhanced transient response demonstrated in simulations

Experimental validation confirms improved stability and performance

Overcomes limitations of small adaptation gain assumption

Abstract

In a recent paper [18] the authors proposed the first solution to the problem of designing a {\em globally exponentially stable} (GES) flux observer for the interior permanent magnet synchronous motor. However, the establishment of the stability proof relies on the assumption that the adaptation gain is sufficiently {\em small} -- a condition that may degrade the quality of the transient behavior. In this paper we propose a new GES flux observer that overcomes this limitation ensuring a high performance behavior. The design relies on the use of a novel theoretical tool -- the generation of a {\em ``virtual" invariant manifold} -- that allows the use of the more advanced Kreisselmeier's regression extension estimator, instead of a simple gradient descent one. We illustrate its superior transient behavior via extensive simulations and {\em experiments}.