Long short-term memory (LSTM) -based neural network model for optimizing composite manufacturing process using autoclave

Abstract Producing high-quality fiber-reinforced composites requires precise temperature control during autoclave curing, as even small variations can lead to defects that compromise strength and reliability. At the same time, manufacturers aim to reduce energy use and shorten curing cycles without sacrificing material performance. To address these challenges, this study develops a data-driven Long Short-Term Memory (LSTM) neural network model capable of forecasting temperature evolution inside the autoclave throughout the curing cycle. The model is trained on time-series temperature data collected from multiple sensing locations, enabling it to learn the spatial and temporal trends that govern heat flow during curing. Data augmentation techniques such as time shifting, scaling, and jittering were applied, helping the model better handle noise and inconsistencies in the dataset. The resulting predictions closely match expected temperature patterns, showing that learning-based models can effectively capture the complex and dynamic thermal behavior within the autoclave. By offering early insight into temperature behavior during curing, the LSTM approach can support better heating control, improve curing consistency, and help reduce overall cycle time. This capability leads to more uniform temperature distribution, fewer unnecessary dwell periods, and higher-quality composite parts. These results show that predictive deep learning can be successfully integrated into autoclave operations, providing a strong foundation for future real-time, adaptive process control in smart composite manufacturing.