Monitoring and diagnostics using state estimation and artificial intelligence

Condition monitoring and diagnostics require detailed information about the performance of the plant so that clues giving an early warning of a failure or a set of failures are noted. State estimation and observer concepts are useful for extracting information from the plant. Artificial intelligence methods can use these information for processing and decision making. This work is concerned with modeling of a plant and generating a set of residual vectors. These vectors are functions of faults. They are therefore are processed by the Inference engine for detection of misbehaviors