A Novel Multilayer N-LMS Adaptive-Filter-Based Control for Synchronization and Power Quality Improvement in Grid-Tied SPV System
Gaurav Modi, Bhim Singh
Abstract
In the power distribution grid, a solar photovoltaic (SPV) system may experience voltage sag, swell, harmonics distortion, unbalance, and dc offset at its point of interconnection. These conditions decline the power quality indices, which are escalated when the SPV system supplies power to unbalanced/distorted loads. In this context, in this article, a multilayer normalized least mean square adaptive-filter-based control is proposed for the grid-tied SPV system, where the sensed grid voltages and load currents are filtered using the proposed method. Then, their direct and quadrature ( <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">dq</i> )-axis components are estimated and used to resolve synchronization issues. The estimated components are immune to negative-sequence components, distortion, low-order odd harmonics (3rd, 5th, 7th, and 9th), and dc offset. The proposed control scheme has a simple structure with two tuning parameters. Moreover, it has no integrator compared with the generalized-integrator-based algorithm and only has three delay units compared with the methods based on the cascaded delay signal cancelation (CDSC). Hence, it is computationally faster than the advanced CDSC and other multilayer-structure-based algorithms. The performance of the proposed method is demonstrated experimentally by implementing it on a laboratory setup of a grid-tied SPV system operating under nonideal voltages and currents.