Model-Free Predictive Current and Speed Control with Ultra-local Model for PMSM Drives Using Exponential High-Gain Extended State Observer

Model-free predictive control of permanent magnet synchronous motors (PMSM) based on ultra-local Model has become a hot research topic in the academic community. Currently, this method mainly focuses on the current loop, and the estimation of disturbances in ultra-local model usually uses an extended state observer. For this issue, this paper proposes model-free predictive current and speed control with ultra-local model for PMSM drives using exponential high-gain extended state observer, abbreviated as MFPCSC-EESO. The MFPCSC-EESO algorithm adopts deadbeat predictive control based on ultra-local model in both the speed and current loops, effectively improving the dynamic response of the speed and stator current. The design of extended state observer is crucial in ultra-local model predictive control. This paper proposes an exponential high gain extended state observer (EESO) and analyzes its construction principle and parameter optimization configuration, further improving the convergence speed and accuracy of disturbance estimation. In addition, the MFPCSC-EESO has been extended to IPMSM with MTPA and field weakening (FW) operation region. The experimental results verified the effectiveness of the proposed method with SPMSM and IPMSM.