Detection of incipient failures using an H2-norm criterion: application to electric point machines

The paper deals with condition monitoring of electric switch machines for railway points. The proposed detection system is based on off-line processing of armature current and voltage sampled as the machine operates a switch of railway tracks. The system, consisting of an algorithm tuned on a model of the machine as both the rails are driven towards their rest positions, allows detection of faults of incremental nature, i.e. those connected to progressive increase of frictional loads (loss of lubrication, deterioration of slide chairs, increasing obstructions, etc). The algorithm realizes a finite impulse response system whose convolution profiles are designed on the basis of an H2-norm criterion which guarantees robustness, particularly with respect to electrical noise.