Organization of Signal Data Sets For Evaluation and Rating of Diagnostic and Prognostic Signal Analysis Techniques

A vast array of signal analysis techniques have been developed over the years to aid in the analysis of vibration, stress wave and other raw sensor output signals. These techniques address the information extraction challenge using many widely varied approaches. With the large variety that exists, determining which solution best fits a given set of requirements can be an expensive and time-consuming process. This process usually includes a large amount of testing to develop the data sets needed to form a common basis for comparison.

Presented within this paper is an organized approach to the collection and cataloging of signal data sets for the express purpose of providing a common basis of comparison, which would greatly simplify the technique evaluation and comparison effort.