Abstract
As the global transition to renewable energy becomes a priority, photovoltaic systems are increasingly important to ensure a sustainable and autonomous power source by exploiting the inexhaustible power of the sun. The power supplied by photovoltaic panels directly depends on climatic conditions, particularly irradiation and temperature. To maximize the energy extracted, it is essential to use a Maximum Power Point Tracking (MPPT) control. Partial shading occurs when certain sections of the photovoltaic array receive reduced irradiation. This phenomenon causes an uneven distribution of solar energy across the panels, leading to changes in their electrical characteristics. However, the performance of MPPT controls can be disrupted by partial shading conditions, complicating optimal operation. This work aims to study two MPPT controls based on the adaptive neuro-fuzzy inference system (ANFIS), each with a different principle, and to analyze and compare their performance in extracting the maximum power available from photovoltaic panels, under uniform and partial shading conditions. The first method combines ANFIS and a fuzzy logic controller, while the second uses ANFIS alone. The comparison will focus on speed, accuracy, and stability, as well as the components required for each method. The results show that both methods perform similarly in accuracy since they can extract almost the same power. However, the second method, which excludes the use of an additional controller, is faster in extracting power with minimal oscillation and reduces the number of components in the photovoltaic system by eliminating the fuzzy controller, thus reducing the system’s complexity.

Key words: ANFIS, MPPT techniques, PARTIAL shading, Photovoltaic Energy, Renewable energy.