Design of a Neural Controller for Single Phase Inverter in Grid Connected Photovoltaic System

Abstract

This study shows a neural network based control strategy of the current injected into a single-phase grid via an inverter. The inverter is supplied by a Photovoltaic Generator (PVG). The optimal control of PVG is ensured by an MPPT algorithm of type P and O (Perturbation-Observation). The synchronization of the inverter with the electrical grid is ensured by a Phase-Locked Loop (PLL) device. The sizing and the modeling of the system components have been presented. A Neural Network Controller (NNC) and a Proportional Integral (PI) controller have been implemented and compared. Obtained results show that the NNC have faster response and lower THD without overshoots.

Adaptive Hybrid MPPT Using Artificial Intelligence for an Autonomous PV System

“…These systems possess the capability to extract information from examples and handle disconnected data. Once trained, the algorithm can quickly perform tasks such as prediction [26,27], optimization [28,29], and system modeling [30,31]. The utilization of ANN eliminates the need to solve intricate mathematical models of systems, and it does not necessitate prior knowledge of the input/output relationships.…”

Design and implementation of a robust ANN-PID corrector to improve high penetrations photovoltaic solar energy connected to the grid

“…The share of renewable energies, reaching 600 MW of total electricity production capacity in 2020, is 13 % hydropower, 13 % solar photovoltaic and 8 % wind power [2]. However, photovoltaic solar energy is able to partially meet current energy needs and overcome the year after year of fossil sources reduction and the emitting aspect of greenhouse gases (GHG) [3,4,5,6,7,8]. It represents an incontestable solution to meet Senegal's energy requirements thanks to its position (5.7 kWh / m 2 per day) and the law No.…”