Multi-objective optimization study on the radiated noise of thin-walled structures in marine diesel engines

Marine diesel engines are the primary source of vibration and noise in ship cabins. As they develop toward higher loads and power densities, these issues can be even worse. Thin-walled structures in the engines further exacerbate the problem. However, experimental methods for vibration and noise optimization are impractical due to the engine’s large size, while simulation-based approaches require extensive computational resources. Thus, developing efficient and intelligent optimization strategies for marine diesel engines’ vibro-acoustic performance remains a challenge. This paper addresses the vibration and noise problem of thin-walled structures in marine diesel engines by proposing an intelligent optimization method that combines the Modal Acoustic Transfer Vector ( MATV ) and Multi-Objective Genetic Algorithm ( MOGA ). First, the Modal Acoustic Transfer Vector ( MATV ) can be obtained via the combination of the Modal Participation Factor ( MPF ) and Acoustic Transfer Vector ( ATV ). MATV helps identify key modal orders affecting radiated noise. Then, an optimization scheme combining anisotropic plate theory and a Multi-Objective Genetic Algorithm ( MOGA ) is proposed. The results show that after optimization, the acoustic contribution of the key modes is reduced by 10% while the overall average Sound Pressure Level ( SPL ) decreases by 12%. At the same time, the vibration intensity is reduced by 10%. These findings validate the effectiveness of the proposed method in low-vibration and low-noise design. This methodology not only offers a novel approach for the intelligent low-vibration and low-noise design of marine diesel engines but can also be extended to the low-vibration and low-noise design processes of other types of main engines and power plants.