Robust three‐voltage‐vector‐based model‐free predictive current control for permanent magnet synchronous motor with ultra‐local model

Abstract

A large current ripple may arise in the finite control set model‐free predictive current control (FCS‐MFPCC) strategy because only one voltage vector can be applied during the whole control period. Therefore, an improved three‐vector‐based model‐free predictive current control (MFPCC) strategy to improve the current tracking performance is proposed by increasing the number of the applied voltage vector. Based on the extreme existence theorem for the function of two variables, the optimal voltage vector can be synthesized. Compared with the one‐vector‐based FCS‐model‐based predicted current control (FCS‐MBPCC) and two‐vector‐based FCS‐MBPCC in the time and frequency domains, the maximum tracking error is decreased by 42.17% and 48.58%, and the total harmonic distortion is decreased by 40.97% and 37.56%, respectively. Besides, compared with the previous three‐vector‐based strategy, the computational time of the proposed strategy is reduced from 57.6 to 45.89 , which is reduced by 20.33%. Furthermore, the switching frequency of the proposed method can be decreased by 70.3% compared with the deadbeat MFPCC. The experimental results in the time and frequency domains demonstrate that the proposed three‐vector‐based MFPCC can not only improve the current tracking performance but also improve robustness against parameter mismatches.

Improved Triple-Voltage-Vector Model-Free Predictive Current Control for Synchronous Reluctance Motor Drives

“…Unlike the ultra-local model-based MFPCC, which relies on a sliding mode observer for calculating current slope through mathematical methods, our approach utilizes current sensors to directly measure current difference. This simplifies the calculation process significantly compared to the duty ratio calculations specified in [55]. Consequently, our method reduces the computational burden relative to the ultra-local model-based MFPCC outlined in [55].…”

“…This simplifies the calculation process significantly compared to the duty ratio calculations specified in [55]. Consequently, our method reduces the computational burden relative to the ultra-local model-based MFPCC outlined in [55]. It was shown in [53]- [57] that the model-free control could be implemented with three vectors of input voltages with high success.…”

“…In addressing issues related to model predictive control systems' performance degradation caused by parameter mismatches and inverter dead-zones, a threevector model-free predictive control (TV-MFPC) strategy is proposed in [22]. Furthermore, [55] delves into a threevector-based MFPCC strategy aimed at enhancing current tracking performance through the application of increased voltage vectors within a single control period. Unlike conventional MFPCC approaches, the method described in [55] computes current variations by multiplying the current slope with an optimal duration derived from an ultra-local model without direct reliance on motor parameters.…”

“…Furthermore, [55] delves into a threevector-based MFPCC strategy aimed at enhancing current tracking performance through the application of increased voltage vectors within a single control period. Unlike conventional MFPCC approaches, the method described in [55] computes current variations by multiplying the current slope with an optimal duration derived from an ultra-local model without direct reliance on motor parameters. Our proposed method offers a distinct advantage in terms of current difference.…”