A qualitative/quantitative representation of signals for supervision of continuous systems

Correct interpretation of measured process data is essential for the supervision of control systems. Qualitative representation of signals based on episodes is presented. A new classification of episodes based on the second derivative sign is performed. Qualitative information associated to each episode is completed with quantitative information. This representation is applied to linear systems responses as an example to distinguish the different dynamics of these systems. Finally, in an application with real signals, its utility and limitations are shown.