Structural damage of metro carbody structures under extreme operating conditions significantly compromises operational safety. Conventional optimization approaches for metro carbody structures are often limited to single load cases and oversimplified methods. To overcome these limitations, this study aims to propose a novel two-stage optimization strategy.
The methodology enables multi-objective optimization analysis incorporating stiffness, static strength, modal performance, mass and fatigue damage, providing engineers with an efficient framework for carbody structural optimization. The study first established a finite element model of the metro carbody and conducted experimental validation. On this basis, a two-stage optimization approach was implemented. In the first stage, topology optimization was performed on the cross-sections of the floor and the main load-bearing roof profiles to determine the optimal material layout. In the second stage, a surrogate model correlating the thickness of highly sensitive components with carbody performance was constructed using a radial basis function neural network. Based on this surrogate model, precise dimensional optimization of the thickness parameters for these highly sensitive components was then carried out employing the improved non-dominated sorting genetic algorithm II.
The optimized carbody structure achieved a 140 kg mass reduction and increased the first-order vertical bending frequency from 9.52 Hz to 10.05 Hz, demonstrating the effectiveness and feasibility of the proposed methodology.
The findings provide valuable insights for optimizing metro carbody performance under complex operational conditions, particularly regarding the balanced improvement of multiple competing performance indicators.
