Fault Detection and Identification for Proprioceptive Sensors Using Analytical Redundancy and Road Signature Determination

A fault detection and identification architecture is presented using analytical redundancy in order to detect faulty comportment from proprioceptive sensors. The three dimensional comportment analysis is then studied in order to create a road signature. This signature is then used to discriminate sensors faults from environmental perturbation. Finally, some strategies are developed to verify the faulty comportment origin, and so ensure the decision process.