Parameter Identification of a PMSG Using a PSO Algorithm Based on Experimental Tests

Abstract

An accurate model for a permanent magnet synchronousgenerator (PMSG) is important for the design ofa high-performance PMSG control system. The performanceof such control systems is influenced by PMSG parametervariations under real operation conditions. In this paper, theelectrical parameters of a PMSG (the phase resistance, the phaseinductance and the rotor permanent magnet (PM) flux linkage)are identified by a particle swarm optimisation (PSO) algorithmbased on experimental tests. The advantages of adopting the PSOalgorithm in this research include easy implementation, a highcomputational efficiency and stable convergence characteristics.For PMSG parameter identification, the normalised root meansquare error (NRMSE) between the measured and simulated datais calculated and minimised using PSO.