MIDPhyNet: Memorized infusion of decomposed physics in neural networks to model dynamic systems

NEUROCOMPUTING

卷:428

Article

Zhang, Zhibo¹  Rai, Rahul^1,2  Chowdhury, Souma³  Doermann, David^4,5 

[1] SUNY Buffalo, Dept Mech & Aerosp Engn, Mfg & Design MAD Lab, 318 Jarvis Hall, Buffalo, NY 14260 USA

[2] Clemson Univ, Int Ctr Automot Res CU ICAR, Geometr Reasoning & Artificial Intelligence GRAIL, 4 Res Dr,346, Greenville, SC 29607 USA

[3] SUNY Buffalo, Dept Mech & Aerosp Engn, Adapt Design Algorithms Models & Syst ADAMS Lab, Buffalo, NY USA

[4] SUNY Buffalo, Dept Comp Sci, Buffalo, NY USA

[5] SUNY Buffalo, Engn & Artificial Intelligence Inst, Buffalo, NY USA

关键词: Hybrid model;    Physics infused machine learning;    Physics guided machine learning;    TCN;    Empirical mode decomposition;   

DOI  :  10.1016/j.neucom.2020.11.042

来源: Elsevier

【 摘 要 】

Integrating simplified or partial physics models with data-driven machine learning models is an emerging concept targeted at facilitating generalizability and extrapolability of complex system behavior predictions. In this paper, we introduce a novel machine learning based fusion model MIDPhyNet that decomposes, memorizes, and integrates first principle physics-based information with data-driven models. In MIDPhyNet the output of partial physics is decomposed into Intrinsic Mode Functions (IMFs), which are then infused to a Memorization Unit to generate embedded vectors. A Prediction Unit synthesizes all of the data to generate prediction results. We test the performance of MIDPhyNet on modeling the behavior of dynamic systems such as an inverted pendulum under wind drag. The results clearly demonstrate the performance benefits of our hybrid architecture over both purely data-driven models and state-of-art hybrid models in terms of generalizability and extrapolability. The MIDPhyNet architecture's superiority is most significant when the models are trained over sparse data sets and in general, MIDPhyNet provides a generic way to explore how physical information can be infused with data driven models. (c) 2020 Elsevier B.V. All rights reserved.

【 授权许可】

Free