Power quality enhancement of dual stage three phase grid integrated SPV system using sequential neural network-based algorithm

Abstract

In this paper, a single-layer sequential neural network (SNN) control algorithm has been proposed for managing the power quality (PQ) in the double-stage three-phase grid-integrated solar photovoltaic (SPV) system. The proposed algorithm efficiently extracts the active and reactive fundamental components of the load current, with a higher convergence rate. The extracted reference currents are matched with the sensed source currents to generate gating pulses for the voltage source converter (VSC). To handle power quality issues in distribution networks such as harmonic imbalances, voltage fluctuations (sags/swells), and the need for active and reactive power compensation, the VSC is controlled very precisely. The control algorithm consists of two-stage processes. In the initial stage, the incremental conductance (INC) based DC-DC converter has been incorporated to extract the peak power from the SPV system. In the later stage SNN-based algorithm is used to control the VSC. A 100 KW solar PV system connected to the utility grid has been designed by using a three-leg VSC in the MATLAB/Simulink environment. The system performance has been evaluated by comparing it to existing techniques. The simulation results comfortably meet the requirements of the IEEE-519–2014 and IEEE 1547–2003 standards for utility grid connection.