Artificial neural network-based reconstruction of temperature fields in two gas oven cavities

Accurate prediction of temperature distributions inside gas oven cavities is essential for improving thermal performance and energy efficiency. In this study, an artificial neural network (ANN) model is developed to reconstruct representative temperature fields using experimentally measured data obtained from two different gas oven configurations. The model is trained using temperature data acquired at multiple measurement locations under various operating conditions, and its performance is evaluated through both qualitative and quantitative comparisons with experimental results. The proposed ANN model accurately reproduces local temperature values, spatial temperature distributions, and global temperature extrema, showing excellent agreement with the experimental data. In addition, the model demonstrates reliable reconstruction under different operating conditions, showing good agreement with experimental measurements for unseen validation cases. These results demonstrate that the proposed approach provides an efficient means of reconstructing complex temperature fields with reduced experimental effort and has strong potential for thermal system analysis and design.