Permanent magnet synchronous motors (PMSMs) have become widely used in high-end equipment manufacturing. Model-free predictive control (MFPC) is implemented to enhance system robustness. With the increasingly stringent requirements for the control performance in high-end applications, the negative effects caused by rotating transform were mitigated, aiming to realize cleaner sampling data and improve model adaptability of the system.

In a study published in IEEE Transactions on Industrial Electronics, a research team led by Prof. WANG Fengxiang from Fujian Institute of Research on the Structure of Matter of the Chinese Academy of Sciences (CAS) designed a novel MFPC strategy using the sinusoidal generalized universal model in a PMSM driving system. 

Researchers analyzed in detail the negative influences and extra harmonics caused by the rotating transform, and designed a sinusoidal generalized universal model in MFPC based on the variable differences, aiming to build the model based on the sinusoidal signals directly. To ensure the optimal adaptability across various operating conditions, all coefficients were estimated using the recurve least square (RLS) algorithm. 

Then, researchers conducted a thorough analysis of the locations of system poles and zeros with different model orders to determine an appropriate compensation gain based on the dominant pole’s location and achieved stability. In cases where the gain exceeds one, a pole shifts into the right-half plane, thereby compromising stability. However, as the gain decreases, the pole migrates towards the left and progressively moves away from the imaginary axis, ensuring the system's stability. 

Besides, researchers showed that this strategy is realized in the continuous control set by the revised causality effectively, enhancing the objective quality through compensation gains. Compared to the traditional time-series-based MFPC, this strategy exhibits superior modeling and prediction accuracy when dealing with sinusoidal causal signals, enhancing the system's robustness. 

This study explores the potential of mitigating the effects of rotational transformations and enhancing model adaptability, culminating in a tailored data-driven model that captures sinusoidal causality.