Enhanced deadbeat predictive current control with novel generalized space vector modulation for five‐level inverter fed permanent magnet synchronous motor drive with reduced torque ripples and less computational burden
Tangirala Isradani,Tejavathu Ramesh
DOI: https://doi.org/10.1002/cta.3943
IF: 2.378
2024-01-26
International Journal of Circuit Theory and Applications
Abstract:Traditional controls like FOC and DTFC are complex due to PI tuning, while MPCC is computationally heavy. Two‐level inverters underperform in high‐power uses compared to multilevel inverters, which, with SVM, add complexity. This work introduces a DBPCC and generalized SVM combo for five‐level inverters, enhancing performance with less computational load. Summary Traditional electric drive control methodologies, such as field oriented control (FOC) and direct torque and flux control (DTFC), predominantly rely on multiple proportional‐integral (PI) controllers that demand complex tuning. While model predictive current control (MPCC) omits PI controllers, it has inherent challenges including high computational burden and tempered dynamics. Although two‐level inverters (2LI) are prevalent in many applications, their performance wanes in high‐power applications like high performance industrial drives and marine propulsions. Multilevel inverters (MLIs) are more efficient, but combined with space vector modulation (SVM), they add computational complexity, particularly at higher levels, limiting drive sampling frequencies and causing torque ripples and harmonics, affecting performance. To address the above drawbacks, this manuscript presents a deadbeat predictive current control of five‐level inverter fed permanent magnet synchronous motor (PMSM) drive with a novel generalized SVM algorithm with reduced computational burden and improved performance. The proposed method is validated through MATLAB/Simulink‐based simulation study and experimental implementation. The results are presented and discussed. The results are compared with DTFC and MPCC where it is observed that the proposed method offers superior performance with reduced computational burden and streamlined implementation.
engineering, electrical & electronic