VIBROTOOL — A Matlab toolbox for change detection and diagnosis in vibration engineering

Change detection and diagnosis (CDD) in industrial processes and equipment are important activities in industry, for maintenance, production cost and safety goals. A set of programs organized as a Matlab™ toolbox, with functions serving the change detection and diagnosis problem is considered. Statistical signal processing methods define the basic kernel meant to build and use programs and functions. Classical methods based on statistical moments of various order and advanced methods based on time-frequency and information analysis are considered. Compared to other toolboxes from the same domain, the VIBROTOOL toolbox promotes mainly blind source separation and signal segmentation procedures for solving the CDD problem. A new advanced procedure based on Hilbert transform and Renyi entropy is considered. The toolbox is designed for vibration signals but can be also used in other domains where change detection is important, as in medicine or environmental changes.