pcNN-SN: a physics-enhanced predictor-corrector neural network algorithm coupled with the discrete ordinates method for neutron transport equations
The algorithm improves coarse-grid solutions for neutron transport calculations by coupling a physics-enhanced predictor-corrector neural network (pcNN) with the discrete ordinates method. It predicts correction factors based on local physical features from transport equations, achieving accuracy comparable to fine grids while reducing computation time by over 60% compared to mesh refinement.