ABSTRACT
The FDD techniques involve three main steps: fault detection, the indication that a fault truly exists; fault isolation or diagnosis, the location of the fault; and fault identification, called fault evaluation according to Isermann (1984), that consists of determining the magnitude of the fault. The methods for fault detection and diagnosis may be classified into two groups: those which do not use a model of the system (modelfree methods), Navarro-Esbrí et al. (2006) and those which do (model-based methods). Here we will focus on the mode-based methods. These methods rely on the concept of analytical redundancy, comparing the model expected performance with the actual measured performance to analyze the presence of faults in the system.
