Share Email Print
cover

Proceedings Paper

Structural damage detection with insufficient data using transfer learning techniques
Author(s): Debejyo Chakraborty; Narayan Kovvali; Bhavana Chakraborty; Antonia Papandreou-Suppappola; Aditi Chattopadhyay
Format Member Price Non-Member Price
PDF $14.40 $18.00

Paper Abstract

The effective detection and classification of damage in complex structures is an important task in the realization of structural health monitoring (SHM) systems. Conventional information processing techniques utilize statistical modeling machinery that requires large amounts of 'training' data which is usually difficult to obtain, leading to compromised system performance under these data-scarce conditions. However, in many SHM scenarios a modest amount of data may be available from a few different but related experiments. In this paper, a new structural damage classification method is proposed that makes use of statistics from related task(s) to improve the classification performance on a data set with limited training examples. The approach is based on the framework of transfer learning (TL) which provides a mechanism for information transfer between related learning tasks. The utility of the proposed method is demonstrated for the classification of fatigue damage in an aluminum lug joint.

Paper Details

Date Published: 18 April 2011
PDF: 9 pages
Proc. SPIE 7981, Sensors and Smart Structures Technologies for Civil, Mechanical, and Aerospace Systems 2011, 798147 (18 April 2011); doi: 10.1117/12.882025
Show Author Affiliations
Debejyo Chakraborty, General Motors Corp. (United States)
Narayan Kovvali, Arizona State Univ. (United States)
Bhavana Chakraborty, Arizona State Univ. (United States)
Antonia Papandreou-Suppappola, Arizona State Univ. (United States)
Aditi Chattopadhyay, Arizona State Univ. (United States)


Published in SPIE Proceedings Vol. 7981:
Sensors and Smart Structures Technologies for Civil, Mechanical, and Aerospace Systems 2011
Masayoshi Tomizuka, Editor(s)

© SPIE. Terms of Use
Back to Top