Email: dh13g23@soton.ac.uk
Doctoral Researcher in Machine Learning at the University of Southampton, specialising in solar irradiance forecasting using computer vision and deep learning. Focusing on leveraging Vision Transformers and hybrid neural architectures to improve the accuracy and transferability of short-term solar forecasting from sky images. This work contributes to advancing renewable energy forecasting under data-scarce and operationally challenging conditions.
David Hamlyn is a doctoral researcher at the University of Southampton, specialising in machine learning and computer vision for solar energy forecasting. His research explores how sky imagery and irradiance measurements can improve the estimation and prediction of solar variability.
His work includes Vision Transformer methods for estimating irradiance with limited training data and GateFormer, a model for intra-hour solar irradiance forecasting. His research interests include multimodal learning, time-series forecasting, data-efficient modelling and evaluating forecasting methods across different sites and atmospheric conditions. His work aims to support the integration of solar generation into energy systems through improved short-term forecasting.
