Performance Improvement of the Grid Connected PV Inverter System with ANN Controller

Abstract

The implementation of the PV system and its integration into the grid has been increased. In this process some power quality issues arise i.e. harmonics, voltage sags / surges, interruptions, flickers, transients, and this is due to non-linear loads, arc furnaces, frequent starting / stopping of electric motors, oscillating loads and interactions of different semiconductor devices. Within these interharmonics is one of the emerging power quality issues in grid-connected photovoltaic (PV) systems. Based on previous case studies and field measurements, evidence of interharmonic emission from maximum power point tracking is one of the leading causes of interharmonics in PV inverters. In this regard, MPPT parameters such as sampling rate and perturbation step size have a strong impact on the interharmonic characteristic of PV system, and to overcome these problems, a mitigating solution has been previously proposed , namely modifications of the MPPT algorithm so as to randomly select the sampling rate between the fast value and the slow value. By implementing this technique with an artificial neural network controller for the control of the inverter. With the proposed method, the voltage perturbations of the DC-link voltage as well as the harmonics of the grid currents are reduced and the performance of the MPPT and PV system has been increased. The performances of the proposed system has been validated on a MATLAB / SIMULINK software environment.