Modeling, simulation and hardware implementation of a bond graph-maximum power point tracker for a photovoltaic panel under partially shaded conditions

Abstract

The electric power generated by a photovoltaic module can be greatly reduced compared with optimal production due to weather conditions and factors such as partial shade. The main defect of a conventional maximum power point tracking control algorithm is its misinterpretation of the location of the maximum power point during a sudden change in climatic conditions because of the existence of several local maxima on the power–voltage characteristic curve. This work presents a preliminary study on the modeling, simulation and implementation of a new algorithm for power output maximization of photovoltaic generators under partially shaded conditions using a bond graph approach. The idea is to use a buck-boost converter and to test experimentally the performance of the proposed algorithm on a real photovoltaic panel. We imposed ten patterns of irradiance on the photovoltaic panel, of which more than half were patterns of partial shading. The proposed controller performed excellently under all shading conditions compared with the classical direct duty cycle technique. The control part and the proposed algorithm were implemented on a microcontroller, and the efficiency of the developed algorithm was demonstrated as a function of the real position of the maximum power point through the results of a simulation performed using Symbols (SYstem Modeling by BOnd graph Language and Simulation) software. The results obtained from the simulation were compared with experimental results obtained from real measurements using a Photowatt PW1650 photovoltaic panel under the same operating conditions and climatic environment.