LED-based solar simulators: A bibliometric review and LSTM-based forecasting approach (2010–2030)

Characterization of photovoltaic technologies and materials requires highly sensitive, spectrally stable, and reproducible light sources. Rigorous evaluation of spectral match, spatial non-uniformity, and temporal instability has established LED-based solar simulators as the current industry and academic standard. Despite the rapid adoption of this technology, existing literature primarily addresses the evolution of LED solar simulators from either a technical or bibliographic perspective, leaving data-driven, future-oriented predictions insufficiently explored. This study examines the evolution of LED-based solar simulators between 2010 and 2025 and generates forecasting scenarios up to 2030 through a comprehensive hybrid methodology integrating bibliometric mapping, critical technical evaluation, and deep learning-based prediction. Within this framework, 333 high-quality publications retrieved from the Web of Science Core Collection were analyzed using VOSviewer and R 4.5.3. The technological evolution of hardware architectures was further examined through an in-depth analysis of 14 high-impact foundational studies. Additionally, a Long Short-Term Memory (LSTM) network was developed to model publication growth trends, achieving a robust coefficient of determination (R^2 = 0.8510) and a Mean Absolute Error (MAE = 1.71). The model predicts a non-linear growth trajectory in academic output, estimating a historical peak of approximately 35 annual publications by 2029. Overall, this study demonstrates the transition of LED solar simulators from experimental laboratory tools to industrial-grade platforms meeting IEC 60904-9 Class AAA standards, driven by modular designs and artificial intelligence integration.