Proposta de um otimizador para rastreamento do ponto de máxima potência de um conjunto de painéis solares conectados em um sistema híbrido de energia

Abstract

This work presents the design and control algorithms of a hybrid power generation system. The system is composed of Back-to-Back converters joined to a solar photovoltaic array, which constitutes the dc-link of the converters. The contribution is the development of five algorithms for tracking the maximum power point (MPPT) of the photovoltaic array. The first algorithm is a modified version of the Perturb and Observe (P&O) algorithm; the second proposed algorithm is based on the gradient method (GM); and the third one is based on optimization of GM (OGM). The last two of the proposed algorithms, are based on a Neural Network, which is combined with the P&O and the OGM algorithms, yielding two hybrid algorithms. The system was developed and simulated using Matlab/Simulink, and the simulation by results are presented for the purpose of evaluating the system and the algorithms performance. The response was evaluated during transient and steady- state conditions, by considering different profiles of load power consumption, irradiance and temperature.