Intelligent multiple sensor early failure warning system for train rotary door operator

The scope of this paper relates to an intelligent multiple sensor early failure warning system. The application is for condition monitoring and incipient-fault reporting of safety-critical single throw mechanical equipment. The train rotary door operator (TRDO) manufactured by Vapor currently in use within London Underground Limited trains is considered as a first case study. Work carried out includes design and development of state-of-the-art control and data acquisition system for Vapor's TRDO. Radial basis networks have been used for on-line system identification and modelling of healthy behaviour of equipment using various sensors. Developing faults are reported based on deviation of any sensor's data from its healthy model. Error data from three sensors monitoring (air flow, air pressure and angular displacement) is fed into two dimensional self organising maps (SOM) to perform the mapping of external sensors error space into the system's internal representational space without human intervention. Different regions are associated to various faulty components/states using database available from fault modes and effects criticality analysis (FMECA) of the TRDO.