Deep neural network enabled corrective source term approach to hybrid analysis and modeling. (February 2022)

• Record Type: Journal Article
• Authors: Blakseth, Sindre Stenen; Rasheed, Adil; Kvamsdal, Trond; San, Omer
• Is Part Of: Neural networks. Volume 146(2022)
• Page Start: 181
• Page End: 199
• Publication Date: 2022-02
• Subjects: Deep neural networks -- Digital twins -- Explainable AI -- Hybrid analysis and modeling -- Physics-based modeling -- Corrective source term approach (CoSTA); Neural computers -- Periodicals; Neural networks (Computer science) -- Periodicals; Neural networks (Neurobiology) -- Periodicals; Nervous System -- Periodicals; Ordinateurs neuronaux -- Périodiques; Réseaux neuronaux (Informatique) -- Périodiques; Réseaux neuronaux (Neurobiologie) -- Périodiques; Neural computers; Neural networks (Computer science); Neural networks (Neurobiology); Periodicals; 006.32
• DOI: 10.1016/j.neunet.2021.11.021
• Languages: English
• ISSNs: 0893-6080

Abstract

Abstract: In this work, we introduce, justify and demonstrate the Corrective Source Term Approach (CoSTA)—a novel approach to Hybrid Analysis and Modeling (HAM). The objective of HAM is to combine physics-based modeling (PBM) and data-driven modeling (DDM) to create generalizable, trustworthy, accurate, computationally efficient and self-evolving models. CoSTA achieves this objective by augmenting the governing equation of a PBM model with a corrective source term generated using a deep neural network. In a series of numerical experiments on one-dimensional heat diffusion, CoSTA is found to outperform comparable DDM and PBM models in terms of accuracy – often reducing predictive errors by several orders of magnitude – while also generalizing better than pure DDM. Due to its flexible but solid theoretical foundation, CoSTA provides a modular framework for leveraging novel developments within both PBM and DDM. Its theoretical foundation also ensures that CoSTA can be used to model any system governed by (deterministic) partial differential equations. Moreover, CoSTA facilitates interpretation of the DNN-generated source term within the context of PBM, which results in improved explainability of the DNN. These factors make CoSTA a potential door-opener for data-driven techniques to enter high-stakes applications previously reserved for pure PBM. Highlights: CoSTA combines differential equations and neural networks CoSTA exploits the best of both the physics-based and data-driven models CoSTA is applied to heat transfer problems with unknown/unmodeled physics CoSTA outperforms both physics-based and data-driven models CoSTA can unlock the full potential of neural networks in high-stakes applications