A Normalized Least Means Absolute Third-Based Adaptive Control Algorithm Design for Grid-Connected PV System

Abstract

This article presents design and real-time control of a single-stage grid-connected photovoltaic (SSGCPV) system. A normalized least means absolute third (NLMAT) algorithm has been employed to design an adaptive controller to achieve effective grid synchronization. The objective is to mitigate harmonics generated by nonlinear load and inject active power to the utility grid with unity power factor (PF). A 20-kW PV system with the above controller has been simulated using MATLAB/Simulink and then verified with a 1.4-kW prototype. A comparison of the proposed controller with least means fourth, variable step size least means square (VSS-LMS), variable leaky least mean square(VLLMS), and NLMAT algorithms has been made. It is found out that although the total harmonic distortion (THD) of the grid current remains within the prescribed limit of the IEEE-519 standard for all the three algorithms, the proposed controller outperforms among all. It provides reduced THD even with unbalanced loading and change in irradiance conditions.