Keywords: train braking system, in-service monitoring, fault reproduction, faultstate space, feature extraction, signed directed graph.

The train braking system is the core part of the security operation of a train. In this paper a fault state-space of braking system based on model-driven method is shown using AMESim which reproduces fourteen types of faults of the braking system, focusing on the mapping between all the faults and their observability, discriminability, severity and involved signals. Further its data-driven approach provides the performance degradation analysis and prediction analysis method of the braking system, which is based on an optimized solution for a test sensor placement scheme in order to extract large data features. A set of state-monitoring system prototypes of the state-monitoring system of in-service braking system is successfully loaded on Guangzhou Metro Line 2 and 3. The simulation and experimental results show that the state-monitoring system developed for the inservice braking system can effectively monitor the braking system faults and provide real-time broadcast. This demostrates that this prototype can effectively predict the performance degradation of key components of the braking system providing an early warning.

purchase the full-text of this paper (price £20)

go to the previous paper
go to the next paper
return to the table of contents
return to the book description
purchase this book (price £65 +P&P)