Adaptive PV array reconfiguration using double deep q-networks for performance improvement under partial shading conditions

This dataset supports the research article titled "Adaptive PV array reconfiguration using double deep q-networks for performance improvement under partial shading conditions". The study addresses the partial shading problem in PV systems by proposing a Double Deep Q-Network (DDQN) based reconfiguration technique. By reducing the overestimation bias inherent in DQN, the proposed method significantly minimizes mismatch losses and improves power output efficiency.
The dataset consists of files containing specifications of the PV module and switching connections, training results, Q-value estimation bias analysis, daily power profiles with energy yields, cumulative energy production data, mismatch loss calculations, and execution time measurements.