A Novel Sensorless Sliding-Mode Control of PMSM with VSI Nonlinearity Compensation

This article presents a novel adaptive full-order sliding-mode observer (AFSMO) tailored for sensorless control of permanent magnet synchronous motors (PMSMs). Additionally, an online compensation scheme for inverter nonlinearity, leveraging the recursive least squares (RLS) algorithm, is proposed. To mitigate the chattering associated with sliding-mode observer (SMO), an adaptive law is introduced to facilitate the extraction of back electromotive forces (EMFs) from the mathematical model of PMSM; Hence, it becomes feasible to assign smaller nonlinear gains to the SMO. Voltage distortion resulting from the nonlinearity characteristics of the inverter induces not only the (6k±1)th harmonics in phase currents but also precipitates an error in the estimation of rotor position. To improve the precision of the estimated voltage distortion, a RLS algorthim is employed to derive the estimated amplitude of the voltage distortion. Ultimately, the proposed method undergoes validation and comparative analysis with conventional approach through experimental procedures.