Comparative performance of dynamic voltage restorer using adaptive control algorithms with optimized error regulator gains

This paper describes the performance of a dynamic voltage restorer (DVR) using various adaptive controls. The implemented control algorithms are; the amplitude adaptive notch filter (AANF), the quasi newton (QN), and the sliding mode observer (SMO) based control algorithms equipped with optimized PI tuning for three leg voltage sources converter-based topology. These control algorithms are assessed and compared using key factors including computational parameters complexity, system response, dynamic and steady state performance etc. The control of a DVR is mainly dependent on the error minimization of the DC voltage and the load terminal voltages, respectively, which directly affects the generation reference load voltage for a DVR. Accordingly, optimization algorithms have been implemented for estimating the PI controllers' gains along with the control algorithms under investigation. The optimization method Harris Hawks Optimization algorithm (HOA) has been implemented for PI tuning for all aforementioned control algorithms. The estimation technique of the PI controller gains using the aforementioned optimization algorithm is elaborated mathematically and pictorially described. A comparative analysis of the performance of the HOA optimization method in estimating the gains of PI controllers of the adaptive control algorithms for three phase DVR are simulated using MATLAB. Moreover, all control algorithms have been developed and tested on the Micro Lab Box based experimental setup of DVR in the laboratory and their performance were compared while considering unbalance for compensation and satisfactory results were obtained.