Phy-Taylor: Physics-Model-Based Deep Neural Networks

by   Yanbing Mao, et al.

Purely data-driven deep neural networks (DNNs) applied to physical engineering systems can infer relations that violate physics laws, thus leading to unexpected consequences. To address this challenge, we propose a physics-model-based DNN framework, called Phy-Taylor, that accelerates learning compliant representations with physical knowledge. The Phy-Taylor framework makes two key contributions; it introduces a new architectural Physics-compatible neural network (PhN), and features a novel compliance mechanism, we call Physics-guided Neural Network Editing. The PhN aims to directly capture nonlinearities inspired by physical quantities, such as kinetic energy, potential energy, electrical power, and aerodynamic drag force. To do so, the PhN augments neural network layers with two key components: (i) monomials of Taylor series expansion of nonlinear functions capturing physical knowledge, and (ii) a suppressor for mitigating the influence of noise. The neural-network editing mechanism further modifies network links and activation functions consistently with physical knowledge. As an extension, we also propose a self-correcting Phy-Taylor framework that introduces two additional capabilities: (i) physics-model-based safety relationship learning, and (ii) automatic output correction when violations of safety occur. Through experiments, we show that (by expressing hard-to-learn nonlinearities directly and by constraining dependencies) Phy-Taylor features considerably fewer parameters, and a remarkably accelerated training process, while offering enhanced model robustness and accuracy.


page 1

page 5

page 7

page 16

page 25


Physical Deep Reinforcement Learning: Safety and Unknown Unknowns

In this paper, we propose the Phy-DRL: a physics-model-regulated deep re...

Physics-based Neural Networks for Shape from Polarization

How should prior knowledge from physics inform a neural network solution...

Supplementation of deep neural networks with simplified physics-based features to increase model prediction accuracy

To improve predictive models for STEM applications, supplemental physics...

Physics-Guided Deep Neural Networks for PowerFlow Analysis

Solving power flow (PF) equations is the basis of power flow analysis, w...

Physics-guided generative adversarial network to learn physical models

This short note describes the concept of guided training of deep neural ...

Enhanced physics-constrained deep neural networks for modeling vanadium redox flow battery

Numerical modeling and simulation have become indispensable tools for ad...

An end-to-end strategy for recovering a free-form potential from a snapshot of stellar coordinates

New large observational surveys such as Gaia are leading us into an era ...

Please sign up or login with your details

Forgot password? Click here to reset