The Information Impulse Function: Detecting Temporal Changes in Structural Response

Intermittent temporal patterns in dynamic response data of nonlinear systems can signify events of high scientific and engineering interest, but confident detection can be elusive. Information theory was investigated as an alternative approach to time history analyses of systems excited by mixed stationary and transient vibration. The resulting technique, called the Information Impulse Function (IIF), operates on time-frequency representations of a signal to detect non-conforming, transient patterns. The IIF was tested with simulated nonlinear response in two different studies with increasing complexity. Laboratory test data was used to investigate method behavior for multiple and long duration events, identify events from background response, and explore the utility of the IIF to discriminate between different transient sources. Despite being a nascent technique, the IIF is a unique and effective addition to the repertoire of signal processing tools used for structural health monitoring, field data analysis, and many other fields of study.