Deep Q-Network Based Global Maximum Power Point Tracking for Partially Shaded PV System

Under the sun insolation in the daytime, the partially shaded effect would easily happen in a photovoltaic (PV) array due to clouds, trees, buildings, etc. To deal with the partially shaded effect, this paper proposes solar global maximum power point tracking (GMPPT) method based on Deep Q-Network. Maximum power point tracking (MPPT) is often used to achieve the maximum power in the PV system. The Perturbation and Observation (P&O) method is one of the most popular MPPT techniques in practice. However, due to the use of fixed step size, the P&O method may cause undesired oscillation around the maximum power point (MPP). With the partially shaded effect, the characteristic P-V curve of a PV array may possess multi-peaks, which often results in tracking of a local maximum, not the expected global maximum. Demonstrated by experiment results, the proposed Deep Q-Network based solar GMPPT method indeed can track the global MPP faster and more precisely without oscillation.