Search

Summary: It is known that the demand for renewable energy systems during recent years has increased significantly, especially for wind and photovoltaic systems. The continuous improvement of the efficiency and performance of these systems is important to their growth, particularly in terms of harvesting the maximum power. Many types of methods and techniques are available for tracking the maximum power point, the most important of which is the concept of artificial intelligence. Artificial intelligence (eg, neural networks and fuzzy neural networks) has great potential to handle complex systems or those that have non‐linear parameters and are influenced by environmental conditions. In this paper, artificial intelligence is used to predict the optimal reference values for wind turbines and photovoltaic panels used in maximum power point tracking (MPPT). Neural networks were used in the wind turbine system, while fuzzy neural networks were used in the photovoltaic systems. The networks were trained in real‐time using an incremental training mode based on the back‐propagation algorithm. The proposed system was designed practically based on the microcontroller and has been tested, and the results are extracted in real‐time. The results have proven their effectiveness and conformity with the required optimal values. [ABSTRACT FROM AUTHOR]

Summary: The performance of the Photovoltaic (PV) system is highly affected by fluctuating climatic conditions. Due to these changing climatic conditions, the irradiance diminishes, and it results in the non‐linearity in voltage‐power (P‐V) curves of the PV system. Therefore, it is one of the reasons for the occurrence of the multiple peaks on P‐V curves instead of a single peak. Therefore, it is challenging to track the global maximum power point tracking (MPPT) under these circumstances with traditional MPPT techniques viz. perturb and observe (P&O) and fuzzy logic control (FLC). Concerning global MPP tracking, this paper proposes a novel MPPT method based on a Rao‐1 optimization algorithm. The proposed algorithm is used to overcome the drawbacks of the classical P&O, FLC, and PSO based MPPT techniques under partical shading conditions (PSCs) and rapidly changing atmospheric conditions viz. slow convergence speed and oscillation around the MPP. The proposed MPPT tracks the global MPP under any abnormal conditions like PSCs. To show the efficiency of the proposed MPPT a comparative assessment is performed with well‐established MPPT methods in the literature. The simulation results witnessed that the proposed MPPT is more superior in improving the PV system performance under normal as well as under PSCs. The proposed MPPT over the PSO technique decreases the tracking time by 72.7% for pattern 1, 29.8% for pattern 2, 66.7% for pattern 3, 61.9% for pattern 5, 27.8% for pattern 6, and improves the tracking efficiency in all shading patterns. [ABSTRACT FROM AUTHOR]