Search for: All records

We report a data-parsimonious machine learning model for short-term forecasting of solar irradiance. The model follows the convolutional neural network – long-short term memory architecture. Its inputs include sky camera images that are reduced to scalar features to meet data transmission constraints. The model focuses on predicting the deviation of irradiance from the persistence of cloudiness (POC) model. Inspired by control theory, a noise signal input is used to capture the presence of unknown and/or unmeasured input variables and is shown to improve model predictions, often considerably. Five years of data from the NREL Solar Radiation Research Laboratory were used to create three rolling train-validate sets and determine the best representations for time, the optimal span of input measurements, and the most impactful model input data (features). For the chosen validation data, the model achieves a mean absolute error of 74.29 W/m2 over a time horizon of up to two hours, compared to a baseline 134.35 W/m2 using the POC model. 

Abstract High-throughput and cost-efficient fabrication of intricate nanopatterns using top-down approaches remains a significant challenge. To overcome this limitation, advancements are required across various domains: patterning techniques, real-time and post-process metrology, data analysis, and, crucially, process control. We review recent progress in continuous, top-down nanomanufacturing, with a particular focus on data-driven process control strategies. We explore existing Machine Learning (ML)-based approaches for implementing key aspects of continuous process control, encompassing high-speed metrology balancing speed and resolution, modeling relationships between process parameters and yield, multimodal data fusion for comprehensive process monitoring, and control law development for real-time process adjustments. To assess the applicability of established control strategies in continuous settings, we compare roll-to-roll (R2R) manufacturing, a paradigmatic continuous multistage process, with the well-established batch-based semiconductor manufacturing. Finally, we outline promising future research directions for achieving high-quality, cost-effective, top-down nanomanufacturing and particularly R2R nanomanufacturing at scale.