Publication

Abstract : The performance of the photovoltaic (PV) array may deteriorate significantly under the partial shading condition (PSC). To reduce these harmful effects, this work introduces a new Internet of Things (IoT)-enabled state-action-reward-state-action (I-SARSA) algorithm that alters the interconnection of traditional total-cross-tied (TCT) PV arrays. The proposed algorithm runs by selecting the random reconfiguring matrix obtained using the IoT platform, following the existing epsilon-greedy policy, and updates its assigned location of PV modules in the form of Q-values. The developed I-SARSA algorithm uses a message queuing telemetry transport (MQTT) communication protocol that enables communication with PV modules and users while PSCs. Subsequently, the redistribution of the shaded PV module takes place dynamically, thereby minimizing mismatch losses and enhancing the global peak. Further, the performance parameters such as mismatch power loss (MPL), fill factor (FF), percentage power loss (%PL), and power enhancement are evaluated for the proposed and various existing approaches of dynamic PV array reconfiguration. The obtained results establish a substantial enhancement across all parameters, with up to 579 W reduced MPL compared to the conventional TCT and a %PL reduction up to 13.467%, hence proving to be a more competent sustainable energy solution. Moreover, the associated parameters of the PV arrays are monitored using the IoT platform. Furthermore, a hardware prototype has also been developed with an ATMEGA 25 board containing ESP 8266 and ICs 74LS138 to evaluate the performance measures of the proposed approach, where the results are found satisfactory.