(369h) Monitoring and Diagnosis of An Experimental System

This presentation will describe efforts put forth for monitoring and diagnosis in an experimental system. The specific system considered is a simple air tank process used for control and modeling exercises. Detailed dynamic modeling will be described. Using a residual-based approach and steady state gain analysis, various faults and their extent can be determined. Additionally, a unscented filter can be used to determine unmeasured parameter values as they change with time.