Abstract
This article proposes a novel sensorless control method for switched reluctance motors (SRMs) operating in medium-to-high speed ranges, based on flux linkage modeling and gradient optimization. The approach first develops a flux linkage interpolation model (FLIM) by characterizing flux linkage properties at both unaligned and aligned positions, establishing a continuous flux linkage model that covers the entire positional domain, significantly simplifying traditional modeling processes. An adaptive gradient optimization method is then designed to estimate the optimal rotor position in real-time by updating interpolation functions based on flux linkage errors. Furthermore, to address phase-fault conditions, a fault-tolerant strategy called faulty-phase flux reconstruction (FPFR) is introduced, which leverages the spatial symmetry of fivephase SRMs to reconstruct flux linkage in faulty phases while maintaining stable system operation. The proposed sensorless control method has been successfully verified through experiments on a five-phase 10/8 SRM.

This publication has 38 references indexed in Scilit: