A multi-convex composition of adaptive LMS filter-based control for dynamic voltage restorer with SSO-optimized PI gains

Chinmay Deshpande,Rajasekharareddy Chilipi,Sabha Raj Arya
DOI: https://doi.org/10.1007/s00202-023-02119-4
IF: 1.63
2024-02-21
Electrical Engineering
Abstract:This research introduces a new approach to enhance the dynamic voltage restorer's performance using a multi-convex adaptive least mean square filter-based control structure. The control structure efficiently extracts the fundamental source voltage component with fast convergence. The main objective of the proposed control scheme is to improve compensation capability and maintain load voltage to be stable with linear and nonlinear loads. It incorporates four specialized least mean square (LMS) sub-filters to overcome fixed step size limitations, enhancing accuracy and tracking of source voltage even under abrupt voltage changes. Each LMS sub-filter excels in robust tracking and significant harmonic reduction, adapting to various step values for improved performance during voltage distortions. The use of distinct mixing parameters for each filter enhances the tracking performance of the control structure. This approach reduces peak overshoot and settling time compared to the traditional LMS method. The proportional–integral (PI) gains of AC bus voltage and DC-link voltage are optimized using salp swarm optimization (SSO) algorithm. System performance validation is done through MATLAB/Simulink simulations that demonstrate compliance with IEEE-519-2014 standards for harmonic spectra.
engineering, electrical & electronic
What problem does this paper attempt to address?