Análise de algoritmos MPPT aplicados a sistemas fotovoltaicos com perturbações de irradiação

Abstract

This work brings an analysis of different approaches used in tracking the maximum power point (MPPT) in photovoltaic systems subject to different conditions of partial shading (situation where the P-V curve of the array presents more than one maximum point), with the objective of determining, among the analyzed approaches, which ones have the best performance. Therefore, computational simulations are performed in Simulink software to evaluate the performance of the following algorithms: Perturb and Observe (PO), Incremental Conductance (IC), Hill Climbing (HC), Particle Swarm Optimization (PSO) and Cuckoo Search (CS). First, simulations are performed with a constant irradiance of 1000 W/m² and a temperature of 25ºC to assess whether the algorithms are capable of tracking the MPP in a situation where there is only one maximum point on the P-V curve. Next, simulations are carried out where the array is subject to three different conditions of partial shading. Analyzing the results, it is observed that all algorithms were able to track the MPP under constant irradiance, although they needed different times to do so. However, under partial shading, only the PO and IC algorithms were able to track the MPP in the three situations, with the HC and CS approaches only tracking the maximum power point in one situation, and the PSO was not able to track the MPP in any situation. In this way, it is concluded that the perturb and observe algorithm obtained the best performance among the others and has a simple structure, a fact that contributes to its being the most used approach in current inverter systems.