Optimization of Resonance Avoidance Design for High-Speed Machines
Year: 2025
Type: article
Abstract: In this study, the design parameters affecting the natural frequency of high-speed machines were optimized. Accurate frequency prediction to avoid resonance is essential during the design stage. Response Surface Method (RSM) and K-Nearest Neighbor algorithm (K-NN) were used as prediction models for optimization. The optimization technique using RSM successfully achieved the target natural frequency, with reliability of the results confirmed by mode matching. Similarly, the K-NN predictions showed similar trends. Accurate prediction was possible with both methods within the search range of the design variables. These findings demonstrate that both approaches can complement each other, providing robust prediction and optimization capability. The results provide useful guidelines for resonance-avoidance design of high-speed rotating systems.
Authors: Jong Yun Kim, Jae Woong Kim, Seong Keol Kim
Language: English
Subfield: Mechanical Engineering
Field: Engineering
Domain: Physical Sciences
Open Access status: diamond
OpenAlex ID: W4415560538