Structural Health Monitoring (SHM) is a process that aims to detect damage in structures
through sensors attached on this structure that capture the data and quantify the damage
in real time. A big number of sensors capturing information in short time intervals can
generate a huge volume of data, that needs to be transmitted and stored. Data compression
then becomes essential to these kinds of system. In the compression process, however,
essential characteristics of the signals generated by the sensors can be lost, and the damage
detection could be compromised. Considering that, this study propose to compare several
combinations of data compression methods and damage detection techniques. To measure
the efficiency of those combinations it is utilized a parametric metric, which quantifies how
good a combination is in a single measure. After comparing the results it was observed
that the algorithm Chebshevy Approximation generate the best results, specially when
combined with Fuzzy-C-Means and K-Means.
|
