Optimization of Anti-Climb Energy-Absorbing Structures in Urban Trains Using RBF Neural Network Modeling and the Multi Objective Genetic Algorithm

Authors

1 University of Science and Technology of Iran, University St., Hengam St., Resalat Square,Tehran

2 University of Science and Technology of Iran, Tehran

10.22044/jsfm.2026.16931.4014

Abstract

Enhancing passenger safety during train collisions requires the design of railway vehicles with efficient energy absorption capacity. Energy-absorbing structures installed at the ends of cars play a vital role in managing impact energy. This research investigates the performance of an energy-absorbing structure comprising a conical tube and an anti-climber plate under eccentric loading. Using the Hammersley sampling method, 100 geometric models with varying design parameters for the conical tube (radius, thickness, and cone angle) were created, and their quasi-static crushing process was analyzed via Finite Element (FE) simulation in Abaqus. Key performance indicators, namely Specific Energy Absorption (SEA) and Maximum Crushing Force (Fmax), were extracted. Due to high computational costs, a high-accuracy (R² = 0.98) surrogate model based on a Radial Basis Function (RBF) network was employed. Multi-objective optimization using the NSGA-II algorithm revealed the Pareto front, illustrating the trade-off between the conflicting objectives. Validation through FE simulation showed a relative difference of less than 2% for both SEA and Fmax in the optimal design selected by the minimum distance method, confirming the accuracy and reliability of the proposed framework. These findings verify the capability and dependability of the proposed framework as an effective guide for the optimal design of energy absorbers in the railway industry.

Keywords

Main Subjects

Journal of Solid and Fluid Mechanics
Volume 16, Issue 1
May and June 2026
Pages 1-14

Files

Share

How to cite

Statistics