Enhanced block-sparse adaptive Bayesian algorithm based control strategy of superconducting magnetic energy storage units for wind farms power ripple minimization

Abstract

This article presents a novel enhanced block-sparse adaptive Bayesian algorithm (EBSABA) to fully control proportional-integral (PI) controllers of superconducting magnetic energy storage (SMES) units. The main goal is to smooth the real power output from grid-tied wind farms (WFs) and enhance its power quality, which represents a significant concern in the industry. In this regard, two WFs are tied to the network and each one is equipped with a SMES unit. The proposed algorithm takes into consideration the effect of actuating error signal and its magnitude to online update the PI controller gains. The proposed control strategy is applied to all power electronic circuit converters. To obtain a realistic work, practical measured data of wind speed that recorded from Hokkaido Island are incorporated into the analyses. Moreover, a three-drive train model represents a turbine model. A practical 10 MW SMES unit is connected at the WFs terminals. The effectiveness of proposed SMES units is verified by comparing its results with those obtained by using the least mean square-PI SMES units and optimal PI SMES units by genetic algorithm under wind speed variability and uncertainty. The real power can be smoothened by more than 10% using the proposed system at some intervals. The validity of the study is tested by the simulation results that are carried out by PSCAD environment. The controlled SMES units can further improve the power quality of WFs.