Deep neural network enabled corrective source term approach to hybrid analysis and modeling
| Autor: | Trond Kvamsdal, Sindre Stenen Blakseth, Adil Rasheed, Omer San |
|---|---|
| Rok vydání: | 2022 |
| Předmět: |
FOS: Computer and information sciences
Computer Science - Machine Learning Mathematical optimization Physics-based modeling Cognitive Neuroscience FOS: Physical sciences Context (language use) Digital twins Machine Learning (cs.LG) Corrective source term approach (CoSTA) Artificial Intelligence Deep neural networks Neural and Evolutionary Computing (cs.NE) Hybrid analysis and modeling Partial differential equation Series (mathematics) Artificial neural network business.industry Physics Computer Science - Neural and Evolutionary Computing Computational Physics (physics.comp-ph) Modular design Term (time) Trustworthiness Explainable AI Heat equation Neural Networks Computer business Physics - Computational Physics |
| Zdroj: | Neural Networks |
| ISSN: | 0893-6080 |
| Popis: | 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. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|
Full Text from ScienceDirect