REDUNDANT SENSOR FAULT DETECTION AND SIGNAL RECONSTRUCTION BASED ON ASSOCIATED NEURAL NETWORKS
-
Graphical Abstract
-
Abstract
This paper discusses the methods of the sensor fault detection and signal reconfiguration using associated neural networks. The first stage of the neural networks compresses redundant informations which are used to filter fault information. The second stage of the neural networks restitutes compressed non-fault information and diagnoses fault using proportion of SPE, then completes signal reconstruction using redundant informations.
-
-