Finite Element Model Calibration with Surrogate Model-Based Bayesian Updating: A Case Study of Motor FEM Model

This paper introduces an advanced methodology for the calibration of Finite Element Models (FEM) utilizing a surrogate model-based Bayesian updating framework. The approach is exemplified through a case study on motor design, where precise FEM calibration is essential for predicting and optimizing motor performance. Traditional calibration techniques are often computationally expensive due to the iterative nature of the simulation process. To mitigate this, the proposed method integrates surrogate models to approximate FEM simulations, significantly reducing the computational burden without sacrificing accuracy. Bayesian updating is then employed to iteratively refine the surrogate model by incorporating new data, thereby enhancing prediction accuracy. This dual approach not only accelerates the calibration process but also ensures a high level of precision, making it highly suitable for complex engineering applications requiring both efficiency and reliability. The case study underscores the effectiveness of this methodology, demonstrating its potential to streamline the design process in motor development and other FEM-dependent engineering fields. The findings suggest that the surrogate model-based Bayesian updating approach achieves robust calibration with significantly fewer simulations, thereby optimizing both time and computational resources.