Technology Outline (Process Description)

The present invention relate to a hybrid hardware technique for solar photovoltaic system with maximum power point (MPP) and faulty panel finding in a single step method. This invention presents a plug and play IV curve tester that deploys periodic sweeps to trace the dynamic IV and PV curves to determine Maximum Power Point (MPP) under normal and fault conditions. The IV curve tester consists of three GaNs (1, 2, 3 &4), a capacitor (C), a discharge resistor (R) and a DC bus capacitor (CB). During the curve tracing mode, the capacitor (C), begins to charge from short circuit to open circuit state by enabling the GaNs (1&4), completing the charging cycle. The voltage and current data points are then measured across and through the capacitor (C). These data points are utilized to trace the IV and PV curves and track the MPP of the SPV array. It is important to note that the supply to the load is temporarily disrupted during this cycle, which isn’t desired. The capacitance (CB) operates through the GaN-2, needs to be designed with reference to the load and should still be able to power the load even if there is a surge/budge in load power requirement.

During discharge mode, it is essential for the capacitor (C) to discharge after successful completion of tracing the IV and PV curves and prepare for the next sweeping cycle. This operation is aided by enabling the discharge resistor (R) through the GaN-3 switch, quickly being able to discharge the capacitor (C) and preparing the tester for the next sweeping cycle. An innovative approach aimed at optimizing the efficiency of solar photovoltaic (PV) systems by replacing the bypass and bypass diodes and switches and leveraging Convolutional Neural Networks for accurately identifying and locating faulty panels. A multi-capacitor-based IV curve tester is employed in parallel with the PV string, enabling periodic generation of IV and PV curves. Within this, the bypass and blocking diodes are replaced with low “ON” state resistance (RDS (ON)) MOSFETs, interconnected in a parallel manner with each panel, functioning as switches that sequentially bypass each panel within the string. The resulting images are then fed to the AI model, which meticulously observes and analyses the curves, determining the exact partially shaded panel. The control panel (communication module) receives this information and remotely bypasses the faulty panel, thereby preventing losses and subsequent physical damage to the solar PV system.