Three‐vector model‐free predictive control for permanent magnet synchronous motor

Abstract

AbstractA three‐vector model‐free predictive control (TV‐MFPC) strategy is proposed to address the problem of model predictive control systems performance degradation caused by parameter mismatch and inverter dead‐zone. Firstly, an ultra‐local model is constructed for parameter mismatch and dead‐zone. Then, to improve the adaptive capability of the observer to parameter variations, sliding mode control is introduced, and an Improved Extended State Observer (IESO) is designed to estimate unknown disturbances in the ultra‐local model and provide feedforward compensation for the current loop. Finally, the proposed strategy is experimentally validated, and the results demonstrate that it reduces dependence on the model, suppresses the phase current distortion caused by dead‐zone, and improves the robustness of the control system.