Crash energy management optimization of high-speed trains by machine learning methods
Shaodong Zheng,Lin Jing,Kai Liu,Zhenhao Yu,Zhao Tang,Kaiyun Wang
DOI: https://doi.org/10.1016/j.ijmecsci.2024.109108
IF: 7.3
2024-02-17
International Journal of Mechanical Sciences
Abstract:With the increasing speed of railway vehicles, the intricacies inherent in train collision systems pose challenges in the rational allocation of energy during collision events. In this study, an efficient strategy for train crash energy management was proposed by integrating machine learning and the multi-objective optimization method. A 3-D finite element model of an eight-marshalling train was established and a train collision database was built by simulating train-to-train collisions. The machine learning methods were used to construct prediction models for the energy absorption in the head car's interface ( E a ) and the standard deviation of the energy absorption in each intermediate car's interfaces ( σ ). The machine learning prediction model served as the fitness function for the multi-objective optimization algorithm, to achieve a maximum E a and minimum σ based on the idea of collision energy management. The sample data of 340 groups were found to be sufficient to construct a machine learning model for energy absorption prediction, and the XGBoost was chosen to predict the collision energy absorption with R 2 of 0.923 for E a and 0.927 for σ , respectively. The optimal alternative of train crash energy management was obtained (i.e., F 1 = 1787.69 kN, F 2 = 2881.38 kN, F 3 = 1596.43 kN, F 4 = 1353.44 kN, F 5 = 1765.68 kN, and F 6 = 1200.64 kN), compared to the traditional configuration of the equivalent values (i.e., 1500 kN). The optimized E a increased by 10.51% and σ decreased by 12.59%, and the main energy absorption interfaces of the intermediate cars changed from the original 6 to 8 interfaces. The optimized train displayed better crashworthiness performance in terms of instantaneous acceleration, living space, and peak interfacial forces. These findings demonstrated that the proposed approach was effective in optimizing train crash energy management.
engineering, mechanical,mechanics