Method and System of Monitoring and Prognostics

Abstract

A neural network learns the operating modes of a system being monitored under normal operating conditions. Anomalies can be automatically detected and learned. A control command can be issued or an alert can be issued in response thereto.

1 - 17 . (canceled)  
   
   
       18 . An apparatus comprising: 
 a plurality of inputs associated with a pre-selected activity;    a neural network which receives signals from the inputs and responsive thereto establishes an expected activity related process; and    network memory wherein a representation of a learned activity can be established.    
   
   
       19 . An apparatus as in  claim 18  wherein the network includes processing circuitry that relates a current activity to a learned activity.  
   
   
       20 . An apparatus as in  claim 19  which includes output circuitry, responsive to the results of relating the current activity to the learned activity, that generates at least one output control signal.  
   
   
       21 . An apparatus as in  claim 18  which includes a plurality of neural networks.  
   
   
       22 . An apparatus as in  claim 21  where members of the plurality sense a common set of inputs.  
   
   
       23 . An apparatus as in  claim 22  where members of the plurality sense different inputs.  
   
   
       24 . A method of monitoring comprising: 
 training members of a plurality of neural networks to recognize a plurality of anomalies; and    sensing outputs from the members of the plurality and responding thereto to generate at least one of a control command or an alert indicating indicium.    
   
   
       25 - 32 . (canceled)