Optimization of wheel profile for flange wear of rail vehicles on worn rails
摘要
Optimizing wheel profiles to align with the characteristics of running lines is critical for reducing wheel wear and enhancing operational safety and ride quality. This study presents a multi-objective optimization framework for wheel profiles, specifically targeting flange wear minimization during interactions with worn rails across varying curve radii. By integrating the RBF neural network with the NSGA-II algorithm, two optimized wheel profiles were generated. Key findings indicate that the optimized profiles achieve significant wheel wear reduction, with pronounced effects under conditions of worn rail profiles and small curve radii. Notably, compared to optimization strategies considering only new rails, the approach accounting for both new and worn rails yields greater flange wear reduction. Furthermore, the optimized profiles improve vehicle safety, where the profile considering both new and worn rails outperforms that considering only new rails in terms of both flange wear reduction and safety enhancement.