A power bond graph based diagnostic algorithm for CCBII brake

• Published in: Chinese Control and Decision Conference pp. 2961 - 2965
• Main Authors: Xiong, Lu, Yang, Yingze, Huang, Zhiwu, Li, Liran
• Format: Conference Proceeding Journal Article
• Language: English
• Published: IEEE 01.05.2016
• Subjects: Algorithms; Bond graphs; Brakes; CCBII brake; Electric power generation; fault diagnosis; Fault trees; Faults; Mathematical models; Monitoring; Parameters; power bond graph; temporal causal graph
• ISSN: 1948-9447
• DOI: 10.1109/CCDC.2016.7531489

As a mainly used Electronic-Controlled Pneumatic brake, CCBII brake is designed with complex operation modes and various components, it is difficult to model and analyze its instantaneous and pneumatic faults. In this paper, a power bond graph based method is proposed for CCBII brake modeling, and a new fault tree technique is used to identify and isolate fault parameters. Firstly, according to different operation modes of CCBII brake, power bond graph is deduced to describe inter-components' connective parameters. Secondly, temporal causal graph based fault tree models are designed with fault set and fault features. Finally, the control deviation of brake pipe reservoir is detected online, and once fault is established, fault characteristics can be obtained by real-time comparisons inner fault set. At last, simulation results validate the proposed algorithm can realize real-time fault detection and isolation for CCBII brake.