Multiparameter Estimation Accuracy Improvement for PMSMs Using a Constriction Coefficient-Based Particle Swarm Optimization

Accurate estimation of electrical parameters is important for high-performance control of permanent magnet synchronous machines (PMSMs). This paper proposes a novel non-invasive multi-parameter estimation method for PMSMs based on a constriction coefficient-based particle swarm optimization (CCPSO) algorithm which accurately estimates the electrical parameters of the machine including its stator winding resistance, permanent magnet (PM) flux linkage, and dq-axis inductances. Voltage source inverter (VSI) non-linearity due to the dead-time effect is compensated, and magnetic saturation is modeled through polynomial functions representing dq-axis inductance variations with currents. Furthermore, a method for decoupling the parameters in PMSM voltage equations is proposed to reduce both the cross-coupling effects between electrical parameters and the computational burden. Three main objective functions are defined and implemented using the CCPSO algorithm to track the parameters and enhance the estimation accuracy. The proposed method is validated through experiments on a laboratory 4.25-kW interior PMSM (IPMSM) over a wide range of operating speeds and loads. Finally, the estimated parameters from the developed CCPSO algorithm are compared to the results from the conventional least squares (LS)-based estimation to justify the higher accuracy of the proposed CCPSO approach.