Improving real-time photovoltaic power delivery through integrated advanced fault localization and particle swarm optimization

This paper presents a detailed methodology for fault detection, localization, and performance optimization within a four-string photovoltaic (PV) array. The initial phase involves the identification and pinpointing of various defects throughout the PV field, with a detection accuracy that can reach 98%. After detecting the faults, the impact on the affected PV generators is quantified, showing a maximum power loss which can reach up to 20–30% depending on the type of the fault and the number of affected cells or generators, this loss can be largely recovered through the proposed optimization strategy. Building on this analysis, the strategy increases overall power output by 10%–25% and maintains voltage balancing across all strings within a deviation of ±10 V. A key component of this approach is the integration of a particle swarm optimization (PSO)-based maximum power point tracking (MPPT) algorithm, which dynamically adjusts to fluctuating operational and environmental conditions. The system ensures smooth transitions between two distinct chopper architectures series or parallel using a balancing system for currents and voltages at the output of each string, approaching the global optimal point with a rate that can be very close to 100%. This mechanism adapts to the specific requirements of each architecture as well as climatic and operational conditions to extract the maximum possible power from the PV generators.