This paper proposes a genetic algorithm (GA)-based method to optimize 1D rotor models, addressing the trade-off between the high accuracy of 3D models and the efficiency of 1D beam models. By optimizing correction parameters for mass, moment of inertia, and Young’s modulus, the discrepancies in amplitude, phase, and gravity-induced displacement are minimized. The GA-optimized 1D model closely matches the 3D model’s dynamic behavior, especially near critical speeds, outperforming the conventional 1D model.

published in ASME Journal of Vibration and Acoustics (International Journal), 2025 Journal Citation Ranking (Q2, Category: Engineering, Mechanical)

This study proposed a AI-based surrogate bearing force prediction model via data-driven FNO, that is 10,000 times faster than Reynolds solver.

published in Mechanical Systems and Signal Processing (International Journal),                2025 Journal Citation Ranking (6/182, Q1 3%, Category: Engineering, Mechanical)

A method to automate the reactor core startup via the reinforcement learning (RL) algorithm is proposed in this paper.

published in Nuclear Engineering and Technology (International Journal),                 2025 Journal Citation Ranking (7/41, Q1 15.9%, Category: Nuclear Science & Tech)

In this study, performance analysis has been carried out using an in-house code for various scroll geometrical parameters such as basic circle radius, initial involute angle, scroll height, gap clearances, and total number of chambers.

published in Tribology and Lubricants (Domestic Journal)