A Novel Model‐Free Adaptive Sliding Mode Control of PMSM Based on Dynamic Linearization

Conventional control methods rely too heavily on a precise permanent magnet synchronous motor (PMSM) model. However, temperature changes and aging can cause the motor parameters to differ from the model, resulting in degraded control system performance. This paper presents an innovative model‐free adaptive sliding mode control (MFASMC) method that integrates dynamic linearization and sliding mode control (SMC). First, the PMSM motion equation is discretized to create a first‐order compact‐form dynamic linearization (CFDL) model. Next, a novel data‐driven SMC method is proposed that integrates the merits of a discrete SMC (DSMC) method and a model‐free adaptive control (MFAC) method based on CFDL. Then, an improved algorithm is developed to estimate the pseudopartial derivative (PPD) for data‐driven control based on the criterion function and considering the input and output (I/O) data of the system. Furthermore, the convergence property of the PPD estimation algorithm and the stability of the MFASMC are comprehensively analyzed. Finally, a comparison of simulation and experimental results with PI and SMC methods demonstrates that the proposed method significantly enhances the robustness of the PMSM system. Notably, the proposed method eliminates dependence on the system model, providing a more adaptable and efficient control solution for practical engineering applications.