Academic

Residual Attention Physics-Informed Neural Networks for Robust Multiphysics Simulation of Steady-State Electrothermal Energy Systems

arXiv:2603.23578v1 Announce Type: new Abstract: Efficient thermal management and precise field prediction are critical for the design of advanced energy systems, including electrohydrodynamic transport, microfluidic energy harvesters, and electrically driven thermal regulators. However, the steady-state simulation of these electrothermal coupled multiphysics systems remains challenging for physics-informed neural computation due to strong nonlinear field coupling, temperature-dependent coefficient variability, and complex interface dynamics. This study proposes a Residual Attention Physics-Informed Neural Network (RA-PINN) framework for the unified solution of coupled velocity, pressure, electric-potential, and temperature fields. By integrating a unified five-field operator formulation with residual-connected feature propagation and attention-guided channel modulation, the proposed architecture effectively captures localized coupling structures and steep gradients. We evaluate RA-PIN

Y
Yuqing Zhou, Ze Tao, Fujun Liu · · 1 min read · 12 views
#cs.LG #physics.comp-ph

arXiv:2603.23578v1 Announce Type: new Abstract: Efficient thermal management and precise field prediction are critical for the design of advanced energy systems, including electrohydrodynamic transport, microfluidic energy harvesters, and electrically driven thermal regulators. However, the steady-state simulation of these electrothermal coupled multiphysics systems remains challenging for physics-informed neural computation due to strong nonlinear field coupling, temperature-dependent coefficient variability, and complex interface dynamics. This study proposes a Residual Attention Physics-Informed Neural Network (RA-PINN) framework for the unified solution of coupled velocity, pressure, electric-potential, and temperature fields. By integrating a unified five-field operator formulation with residual-connected feature propagation and attention-guided channel modulation, the proposed architecture effectively captures localized coupling structures and steep gradients. We evaluate RA-PINN across four representative energy-relevant benchmarks: constant-coefficient coupling, indirect pressure-gauge constraints, temperature-dependent transport, and oblique-interface consistency. Comparative analysis against Pure-MLP, LSTM-PINN, and pLSTM-PINN demonstrates that RA-PINN achieves superior accuracy, yielding the lowest MSE, RMSE, and relative $L_2$ errors across all scenarios. Notably, RA-PINN maintains high structural fidelity in interface-dominated and variable-coefficient settings where conventional PINN backbones often fail. These results establish RA-PINN as a robust and accurate computational framework for the high-fidelity modeling and optimization of complex electrothermal multiphysics in sustainable energy applications.

Executive Summary

This article introduces a Residual Attention Physics-Informed Neural Network (RA-PINN) framework for robust multiphysics simulation of steady-state electrothermal energy systems. The proposed architecture effectively captures localized coupling structures and steep gradients, achieving superior accuracy compared to existing methods in four representative energy-relevant benchmarks. The results demonstrate RA-PINN's potential as a robust and accurate computational framework for high-fidelity modeling and optimization of complex electrothermal multiphysics in sustainable energy applications. The study's findings and methodology contribute to the development of more efficient and precise energy systems, with significant implications for the design of advanced energy systems.

Key Points

  • ▸ RA-PINN framework integrates a unified five-field operator formulation with residual-connected feature propagation and attention-guided channel modulation.
  • ▸ The proposed architecture effectively captures localized coupling structures and steep gradients.
  • ▸ RA-PINN achieves superior accuracy compared to existing methods in four representative energy-relevant benchmarks.

Merits

Robust and Accurate Simulation

RA-PINN framework provides a robust and accurate computational framework for high-fidelity modeling and optimization of complex electrothermal multiphysics in sustainable energy applications.

Improved Efficiency and Precision

The proposed architecture enables efficient thermal management and precise field prediction in advanced energy systems.

Generalizability and Flexibility

RA-PINN framework can be applied to various energy-relevant scenarios, including constant-coefficient coupling, indirect pressure-gauge constraints, temperature-dependent transport, and oblique-interface consistency.

Demerits

Limited Scalability

The complexity and computational requirements of the RA-PINN framework may limit its scalability to larger and more complex energy systems.

Dependence on High-Quality Training Data

The performance of RA-PINN framework relies heavily on the availability and quality of training data, which can be challenging to obtain and preprocess.

Expert Commentary

This article makes a significant contribution to the field of physics-informed neural networks and electrothermal energy systems. The proposed RA-PINN framework demonstrates superior accuracy and robustness compared to existing methods, and its potential applications in sustainable energy systems are vast. However, the study's limitations, such as limited scalability and dependence on high-quality training data, should be addressed in future research. Additionally, the study's findings and methodology can be extended to other areas of research, such as materials science and engineering.

Recommendations

  • ✓ Future research should focus on addressing the scalability and data quality limitations of the RA-PINN framework.
  • ✓ The proposed framework should be applied to real-world energy systems to evaluate its practicality and potential for real-world impact.

Sources

Original: arXiv - cs.LG

Related Articles