A Method Selection Framework for Operating-Profile Assessment and Sensitivity Analysis of DC/AC Inverters in Photovoltaic Plants

This research proposes a method selection framework for operating-profile assessment and sensitivity analysis of DC/AC inverters in photovoltaic plants. The inverter operating profile is defined as a set of input factors, output indicators, and operational constraints, including solar irradiance, temperature, DC and AC electrical variables, grid conditions, efficiency, power factor, harmonic distortion, MPPT behavior, thermal loading, clipping events, and derating conditions. The main focus is the systematic selection of appropriate assessment and sensitivity analysis methods according to the objective of the study, data availability, required accuracy, interpretability, computational complexity, and contribution to sustainable control and monitoring. Empirical, statistical, model-based, stochastic, machine learning, and hybrid approaches are considered, together with local and global sensitivity analysis techniques such as local derivatives, Morris screening, and Sobol indices. A multi-criteria scoring scheme and a scenario-based decision matrix are proposed to support the selection of suitable methods under different operating and monitoring conditions. The proposed framework can assist photovoltaic plant operators in diagnostics, predictive maintenance, control tuning, loss reduction, and lifetime extension of DC/AC inverters. The results also support the integration of sensitivity-driven monitoring into SCADA-based supervision and digital-twin-oriented decision-making workflows.