Proceedings of the

The 33rd European Safety and Reliability Conference (ESREL 2023)
3 – 8 September 2023, Southampton, UK

Dynamic Risk Assessment of Train Brake System Failures Considering the Component Degradation

Xiaoliang Yin1,a, Jinduo Xing1,b, Jun Zhang1,c and Chen Juan2

1School of Mechanical-Electronic and Vehicle Engineering, Beijing University of Civil Engineering and Architecture, China.

2School of Mechanical Engineering and Automation, Laboratory of Aerospace Servo Actuation and Transmission, Beihang University, China.


Train brake system plays a vital role in train safety. In this paper, a hybrid model is proposed to evaluate the train brake system failure risk. The hybrid model, which combines fault trees with Bayesian networks, has a good logical structure and probabilistic reasoning ability. The fault tree model is used to identify the risk influencing factors in the brake system, while the failure dynamic nature is captured by the Dynamic Bayesian network. In particular, we evaluate the degradation of four common failures, insufficient braking, brake test failure, braking relieve failure and wheel lock. The risk influencing factors of the brake system and their relevance are also identified. A model based on fault tree and Dynamic Bayesian network for the train brake system is developed. The model can capture the spatial variability of parameters and simulates the evolution of brake faults in time and space. The information is used to perform sensitivity analysis and diagnostic inference on the model.

Keywords: Dynamic Risk Assessment (DRA), Train brake system, Bayesian Network (BN), Probabilistic Risk Assessment (PRA), Fault Tree (FT).

Download PDF