Physical Knowledge

Physical knowledge integration into machine learning models is a burgeoning field aiming to improve the efficiency, accuracy, and generalizability of data-driven predictions in various domains. Current research focuses on incorporating partial physical knowledge into neural networks, particularly through physics-informed neural networks (PINNs) and hybrid approaches combining data-driven and physics-based methods. These advancements address challenges like limited data availability and high computational costs, leading to improved performance in applications such as process systems modeling, reinforcement learning, and dynamic process operations. The resulting models offer a more robust and efficient way to analyze complex systems where complete physical descriptions are unavailable or computationally intractable.

Papers

July 7, 2024

PAPM: A Physics-aware Proxy Model for Process Systems
Pengwei Liu, Zhongkai Hao, Xingyu Ren, Hangjie Yuan, Jiayang Ren, Dong Ni
Physic Aware Physical Prior Proxy Model Process System Physical Knowledge

July 2, 2024

Physics-Informed Model and Hybrid Planning for Efficient Dyna-Style Reinforcement Learning
Zakariae El Asri, Olivier Sigaud, Nicolas Thome
Reinforcement Learning Efficient Reinforcement Learning Hierarchical Planning Physic Aware Model Free Policy Hybrid Planning Physical Knowledge

June 3, 2024

Physics-Informed Neural Networks for Dynamic Process Operations with Limited Physical Knowledge and Data
Mehmet Velioglu, Song Zhai, Sophia Rupprecht, Alexander Mitsos, Andreas Jupke, Manuel Dahmen
Neural Network Raw Data Physic Informed Neural Network Process Model Mechanistic Model Mechanistic Study Mechanistic Understanding Dynamic Process Physical Knowledge

October 30, 2023

Disentangled Counterfactual Learning for Physical Audiovisual Commonsense Reasoning
Changsheng Lv, Shuai Zhang, Yapeng Tian, Mengshi Qi, Huadong Ma
Counterfactual Learning Visual Commonsense Reasoning Physical Commonsense Physical Knowledge

April 27, 2023

Some of the variables, some of the parameters, some of the times, with some physics known: Identification with partial information
Saurabh Malani, Tom S. Bertalan, Tianqi Cui, Jose L. Avalos, Michael Betenbaugh, Ioannis G. Kevrekidis
Dynamical System Person Identification Many Parameter Theoretical Physic Variable Model Incomplete Information Experimental Data Physical Knowledge

Physical Knowledge

Papers

PAPM: A Physics-aware Proxy Model for Process Systems

Physics-Informed Model and Hybrid Planning for Efficient Dyna-Style Reinforcement Learning

Physics-Informed Neural Networks for Dynamic Process Operations with Limited Physical Knowledge and Data

Disentangled Counterfactual Learning for Physical Audiovisual Commonsense Reasoning

Some of the variables, some of the parameters, some of the times, with some physics known: Identification with partial information