Learning to Reconstruct Temperature Field from Sparse Observations with Implicit Physics Priors

TL;DR

This paper introduces IPTR, a physics-guided deep learning framework that reconstructs temperature fields from sparse data by leveraging reference simulations, significantly improving accuracy and generalization over existing methods.

Contribution

The paper proposes a novel dual-branch neural network architecture that integrates physics priors from simulations with sparse observations for temperature field reconstruction.

Findings

IPTR outperforms existing methods in accuracy across various settings.

The framework demonstrates strong generalization to different conditions.

Extensive experiments validate the effectiveness of the physics-guided approach.

Abstract

Accurate reconstruction of temperature field of heat-source systems (TFR-HSS) is crucial for thermal monitoring and reliability assessment in engineering applications such as electronic devices and aerospace structures. However, the high cost of measurement acquisition and the substantial distributional shifts in temperature field across varying conditions present significant challenges for developing reconstruction models with robust generalization capabilities. Existing DNNs-based methods typically formulate TFR-HSS as a one-to-one regression problem based solely on target sparse measurements, without effectively leveraging reference simulation data that implicitly encode thermal knowledge. To address this limitation, we propose IPTR, an implicit physics-guided temperature field reconstruction framework that introduces sparse monitoring-temperature field pair from reference simulations as priors to enrich physical understanding. To integrate both reference and target information, we design a dual physics embedding module consisting of two complementary branches: an implicit physics-guided branch employing cross-attention to distill latent physics from the reference data, and an auxiliary encoding branch based on Fourier layers to capture the spatial characteristics of the target observation. The fused representation is then decoded to reconstruct the full temperature field. Extensive experiments under single-condition, multi-condition, and few-shot settings demonstrate that IPTR consistently outperforms existing methods, achieving state-of-the-art reconstruction accuracy and strong generalization capability.