A Sequence-to-Sequence Multi-Fidelity Surrogate Modeling Approach for the Reflow Soldering Process

International Conference on Robotics, Intelligent Control and Artificial Intelligence (RICAI), pp. 822–826

Abstract

Reflow soldering is a critical process in Surface Mount Technology (SMT), directly influencing the performance and reliability of printed circuit boards (PCBs). Precise temperature control during reflow is essential to ensure product quality, yet traditional optimization through physical trials or high-fidelity simulations such as Computational Fluid Dynamics (CFD) is often computationally expensive and time-consuming. To address this challenge, this study proposes a multi-fidelity surrogate modeling approach based on a Sequence-to-Sequence (Seq2Seq) architecture with Long Short-Term Memory (LSTM) networks. The model integrates low-fidelity data derived from convective heat transfer partial differential equations and a limited set of high-fidelity data from finite element CFD simulations. This combination enables efficient and accurate prediction of thermal behavior during reflow soldering. Experimental validation on industrial reflow oven data demonstrates that the proposed method achieves high predictive accuracy even with limited high-fidelity samples. This approach offers a practical solution for rapid and precise thermal process optimization in electronic manufacturing.

Authors 4

1. Xiaohan Wei

   Xi'an Jiaotong University

   Affiliation as printed

   School of Mechanical Engineering, Xi'an Jiaotong University,Xi'an,China

2. Qing Zhang

   Xi'an Jiaotong University

   Affiliation as printed

   School of Instrument Science and Technology, Xi'an Jiaotong University,Xi'an,China

3. Thomas Bäck Aachen Leiden Institute of Advanced Computer Science

   Leiden University

   Affiliation as printed

   Leiden Institute of Advanced Computer Science, Leiden University,Leiden,Netherlands

4. Hao Wang Aachen Leiden Institute of Advanced Computer Science

   Leiden University

   Affiliation as printed

   Leiden Institute of Advanced Computer Science, Leiden University,Leiden,Netherlands