Abstract

Photovoltaic (PV) panels are the most important component in solar energy systems, converting solar energy from sunlight into electricity for production and utility. Monitoring and management processes for solar power systems require ideal operating values ​​of photovoltaic panels as a prerequisite for analyzing, detecting, or classifying fault conditions methodologies. These parameters can be calculated using equivalent electronic circuit analysis or by applying machine learning models trained from previously collected operating data. However, traditional circuit analysis methods require a large amount of computation using iterative search algorithms, while machine learning models are limited by the conditions of training data collection and volume of data required. This study presents a process for constructing an artificial neural network (ANN) using training data built from electronic circuit modeling calculations equivalent to the Newton-Raphson method, aiming to create a method to generate high-accuracy, high-speed reference model for photovoltaic panels to meet real-time monitoring requirements