Predictive Current Control for PMSM Systems Using Extended Sliding Mode Observer with Hurwitz-Based Power Reaching Law

An extended sliding mode observer (ESMO) with Hurwitz-based power reaching law (HPRL) is proposed in this article to improve the performance and the robustness of predictive current control (PCC). First, extended state-space realizations of permanent magnet synchronous motor (PMSM) stator currents mathematical model are introduced by incorporating parameter mismatches as lumped disturbance and state variables. ESMO with a multidimensional sliding mode surface is designed to estimate the stator current and the lumped disturbance. Second, based on the Hurwitz criterion, HPRL is proposed to eliminate the uncertain disturbance calculation of classical power reaching law. Compared with exponential reaching law, HPRL adopted in ESMO achieves high convergence rate and settles chattering. Third, the predictive current discrete model and the predictive lumped disturbance model are presented by using ESMO-HPRL for PCC cost function evaluation. The experimental results validate high robustness and excellent steady and dynamic performance of PCC-ESMO-HPRL in PMSM systems compared with conventional method.