Novel detection soft-computing algorithm for structural health monitoring system

• Main Author: Ng, Kian Theng
• Format: Thesis
• Language: English
• Published: 2013
• Online Access: http://psasir.upm.edu.my/id/eprint/47582/1/FK%202013%2059R.pdf; http://psasir.upm.edu.my/id/eprint/47582/
• Institution: Universiti Putra Malaysia

In view of the criticality of structural integrity, there are many approaches and Structural Health Monitoring (SHM) techniques developed to detect damage or crack in engineering structures. However, there is a lack of attention on developing SHM software for the damage detection in the market. This research aims to develop Novitect, a soft computing platform system for SHM. Novitect is developed using MATLAB 2008-Graphical User Interface (GUIDE) module and designed to formulate novelty detection method employing outlier analysis (OA) for multivariate problem. This system allows users to calculate the novelty index and to display the novelty index plot by processing the data which are in either .MAT or .XLSX format. For this work, the data was based on the acquired waveform response via a smart sensor, piezoceramic transducer (PZT). In the first phase of the designing the Novitect software, the validation of the developed soft computing platform was carried out using acquired data from a hollow cylinder-like structure available from the existing published work. The effectiveness of the developed Novitect was done by executing experiment work on aircraft spoiler which consist of two structural conditions; undamaged and damaged. It was discovered that Novitect is capable to be served as the in-service SHM maintenance tool.