A methodology for the identification of the critical components of the electrical distribution network of cern’s large hadron collider

We consider the propagation of disturbances in the electrical distribution network of CERN’s (European Organization for Nuclear Research) Large Hadron Collider (LHC), which is a complex system made of a large number of mutually interconnected and interdependent components. The objective of this work is the identification of the components most critical in the determination of the system operating/failed state. Given the complexity of the system, the identification is to be based on the operational data collected from the monitoring systems. The critical components are sought as those whose condition monitoring signals are most correlated to the system operating/failed state. The method of the identification is based on the use of the Relief feature selection technique to rank the monitoring signals according to their importance with respect to the classification of the system operating/failed state. The criticality of the identified signals and the corresponding components has been validated by the system technical responsibles.